Complete Technical Paper Content โ Ai Financial Planning
1. Executive Abstract, Problem Formulation & Background
Awesome AI in Finance There are millions of trades made in the global financial market every day. Data grows very quickly and people are hard to understand. With the power of the latest artificial intelligence research, people analyze & trade automatically and intelligently. This list contains the research, tools and code that people use to beat the market. [ ไธญๆ่ตๆบ ] Contents Agents LLMs Skills MCP Servers Market Data Trading Execution Research & Analysis Papers Courses & Books & Blogs Strategies & Research Time Series Data Portfolio Management High Frequency Trading Event Drive Crypto Currencies Strategies Technical Analysis Lottery & Gamble Arbitrage Data Sources Research Tools Trading System TA Lib Exchange API Articles Others Agents ๐๐ nofx - A multi-exchange Al trading platform with multi-Ai competition self-evolution, and real-time dashboard. TradingAgents - Multi-Agents LLM Financial Trading Framework. ๐ FinRobot - An Open-Source AI Agent Platform for Financial Analysis using LLMs. AgentFund - Decentralized crowdfunding platform for AI agents with milestone-based escrow on Base blockchain. ๐ ATLAS - Self-improving AI trading system with 25 agents, Karpathy-style autoresearch, Darwinian selection, autonomous agent spawning, and multi-cohort meta-weighting. InvicTrade - AI-powered trading signals with 74% historical win rate, combining strategies from legendary investors using multi-model AI intelligence. OpenFinClaw - AI-native one-person hedge fund platform. Expert agent teams turn natural language into quant strategies in 60s. Multi-market (US/HK/CN/Crypto), self-evolving strategy pipeline with community leaderboard. Vibe-Trading - Multi-agent finance research workspace for strategy generation, backtests, portfolio analysis, and research insights. TraceArena - Open-source runtime for auditable multi-agent investment evaluation with evidence-linked actions, deterministic simulated settlement, and reproducible replay; no brokerage connection. ProfitPlay Agent Arena - Open prediction market arena where AI agents compete in real-time BTC/ETH/SOL prediction games. Python and Node.js SDKs, 9 live markets, REST + WebSocket APIs. oracle3 - Prediction-market trading agent for Kalshi, Polymarket, and Solana DFlow, with Wang Transform pricing and arbitrage strategies. stock-analysis - Evidence-driven market recap CLI for AI agents, producing Markdown reports and JSON Evidence Packs for A/HK/US stocks, funds, and portfolios. Cod3x - No-code platform for building multi-agent trading strategies, with chart-drawing agents, event-driven automations, and full execution transparency. Pineify - AI-assisted trading toolkit with coding agents for Pine Script, MQL5, and cTrader, plus financial research, strategy optimization, and backtest analysis. Nova-TradingAgent - Self-hosted A-share research workbench built on TradingAgents, with a web interface and no brokerage connection by default. OpenCandle - Open-source financial research agent with real-time market data, local portfolios and watchlists, and no trade execution. LLMs ๐๐๐ Nof1 - Benchmark designed to measure AI's investing abilities. Each model is given $10,000 of real money, in real markets, with identical prompts and input data. ๐ AI Hedge Fund - Explore the use of AI to make trading decisions. ๐๐ MarS - A Financial Market Simulation Engine Powered by Generative Foundation Model. ๐๐ Financial Statement Analysis with Large Language Models - GPT-4 can outperform professional financial analysts in predicting future earnings changes, generating useful narrative insights, and resulting in superior trading strategies with higher Sharpe ratios and alphas, thereby suggesting a potential central role for LLMs in financial decision-making. FinRpt - Dataset, Evaluation System and LLM-based Multi-agent Framework for Equity Research Report Generation. PIXIU - An open-source resource providing a financial large language model, a dataset with 136K instruction samples, and a comprehensive evaluation benchmark. FinGPT - Provides a playground for all people interested in LLMs and NLP in Finance. MACD + RSI + ADX Strategy (ChatGPT-powered) by TradeSmart - Asked ChatGPT on which indicators are the most popular for trading. We used all of the recommendations given. A ChatGPT trading algorithm delivered 500% returns in stock market. My breakdown on what this means for hedge funds and retail investors Use chatgpt to adjust strategy parameters Hands-on LLMs: Train and Deploy a Real-time Financial Advisor - Train and deploy a real-time financial advisor chatbot with Falcon 7B and CometLLM. ChatGPT Strategy by OctoBot - Use ChatGPT to determine which cryptocurrency to trade based on technical indicators. LLMs Meet Finance - A three-stage fine-tuning pipeline (SFT โ DPO โ synthetic-data RL) that adapts Qwen2.5 and DeepSeek-R1 to financial tasks on the Open FinLLM Leaderboard, with findings on cross-task transfer and data scaling laws in finance. MMESGBench - First multimodal benchmark for understanding and complex reasoning over real-world ESG reports; 933 expert-validated QA pairs across 45 documents with text/table/chart evidence and single-page, cross-page, and unanswerable questions (ACM MM 2025). ESGenius - The first benchmark for evaluating LLMs on Environmental, Social, and Governance (ESG) and sustainability knowledge; 1,136 expert-validated multiple-choice questions plus a curated source corpus, evaluating 50 LLMs under zero-shot and RAG (EMNLP 2025). Skills XVARY Stock Research โ Claude Code skill for public SEC EDGAR + market data: /analyze , /score , /compare . MIT. CFA Institute Bias Detection - Claude skill for bias detection in investment analysis. Apache 2.0. Ethical Capital Skills - Claude skills for investment research, screening, compliance, and marketing workflows. Trading Ledger - Claude skill for trading journaling: captures thesis, plan, and emotion at entry into the user's own Notion database, with weekly reviews that grade decisions rather than P&L. MIT. Quant Research - Agent skill for backtest validation using parameter stability, selection-bias checks, and walk-forward evaluation. Market Brief - Agent skill for source-linked funding and liquidity briefs, preserving observation dates and missing-data states. AI Trader Team - Claude Code skills and agents for investment research, with deterministic position-sizing, portfolio-risk, and walk-forward backtesting tools. Bargo Congress API - Agent skill for querying U.S. congressional trade disclosures by ticker, member, chamber, transaction type, and date. financeskills - Agent skills for financial reporting, audit, reconciliation, journal entries, and accounting workflows referencing IFRS and GAAP. MCP Servers Open-source Model Context Protocol servers that connect AI assistants (Claude, Cursor, VS Code Copilot, etc.) directly to financial data and trading APIs. Market Data atilaahmettaner/tradingview-mcp - 30+ tools for real-time TradingView market data, technical analysis, screeners, and backtesting across stocks, crypto, forex, and futures. dgunning/edgartools - 11 MCP tools for SEC EDGAR โ every filing type (10-K, 10-Q, 8-K, S-1), no API key required; 1,800+ Smithery installs. financial-datasets/mcp-server - Structured access to income statements, balance sheets, cash flows, stock prices, and market news via the Financial Datasets API. guangxiangdebizi/FinanceMCP - Integrates Tushare (Chinese A-shares, funds, bonds, macro) and Binance (crypto) for LLM financial data access. TickDB/tickdb-unified-realtime-marketdata-api - 13-tool unified real-time and historical market data across Forex, US/HK/CN stocks, and crypto. massive-com/mcp_massive - Polygon.io professional-grade real-time and historical data for stocks, forex, crypto, and options. Alex2Yang97/yahoo-finance-mcp - Yahoo Finance MCP: stock quotes, financials, options chains, and market news. zwldarren/akshare-one-mcp - AKShare API for Chinese stock market: historical prices, real-time quotes, financial reports, and news. alphavantage/alpha_vantage_mcp - Official Alpha Vantage MCP server: real-time and historical data for stocks, forex, crypto, options, and ETFs. imbenrabi/Financial-Modeling-Prep-MCP-Server - 250+ Financial Modeling Prep API tools: financials, technical indicators, insider trading, SEC filings, earnings, and crypto. kukapay/crypto-indicators-mcp - Cryptocurrency technical analysis indicators (MACD, RSI, Bollinger Bands) via CCXT for AI trading agents. stefanoamorelli/fred-mcp-server - Federal Reserve Economic Data (FRED) MCP server: access 800,000+ macroeconomic time series. livetennisapi/livetennisapi-mcp - Open-source MCP server for tennis scores, fixtures, rankings, and win probabilities for prediction-market research. API plan limits apply. Trading Execution alpacahq/alpaca-mcp-server - Official Alpaca MCP server: natural-language trading of US stocks, ETFs, options, and crypto. krakenfx/kraken-cli - Official Kraken AI-native CLI with built-in MCP support for spot trading, portfolio management, and market data. ariadng/metatrader-mcp-server - MCP bridge to MetaTrader 5 for natural-language trade execution and live position tracking across forex and futures. mcpdotdirect/evm-mcp-server - 22-tool MCP server for 60+ EVM-compatible networks: token balances, smart contracts, and DeFi operations. okx/agent-trade-kit - Official OKX MCP server: AI agent trading of spot, perpetuals, and futures with full order management. rcontesti/IB_MCP - Interactive Brokers TWS/Gateway MCP server for live trading and market access across 150 markets. QuantConnect/mcp-server - Official QuantConnect MCP server: write Python trading strategies, run cloud backtests, and deploy live algorithms. koreainvestment/open-trading-api - Official Korea Investment & Securities Open API with MCP and LLM integration for Korean equity market trading. Research & Analysis mnemox-ai/tradememory-protocol - Open-source memory MCP for AI trading agents: three-layer architecture (raw memory โ pattern reflection โ strategy evolution). wshobson/maverick-mcp - Personal stock analysis MCP server: technical indicators, earnings calendars, options flow, and insider trades. stefanoamorelli/sec-edgar-mcp - Direct SEC EDGAR access for AI assistants: 10-K, 10-Q, 8-K filings, and insider trading data. heubme2020/datasinking - Open-source MCP server for Asian financial reports as Markdown, with chapter-level access. Requires a DataSinking API key; usage limits apply. Papers The Theory of Speculation L. Bachelier, 1900 - The influences which determine the movements of the Stock Exchange are. Brownian Motion in the Stock Market Osborne, 1959 - The common-stock prices can be regarded as an ensemble of decisions in statistical equilibrium. An Investigation into the Use of Reinforcement Learning Techniques within the Algorithmic Trading Domain, 2015 A Deep Reinforcement Learning Framework for the Financial Portfolio Management Problem Reinforcement Learning for Trading, 1994 Dragon-Kings, Black Swans and the Prediction of Crises Didier Sornette - The power laws in the distributions of event sizes under a broad range of conditions in a large variety of systems. Financial Trading as a Game: A Deep Reinforcement Learning Approach - Deep reinforcement learning provides a framework toward end-to-end training of such trading agent. Machine Learning for Trading - With an appropriate choice of the reward function, reinforcement learning techniques can successfully handle the risk-averse case. Ten Financial Applications of Machine Learning, 2018 - Slides review few important financial ML applications. FinRL: A Deep Reinforcement Learning Library for Automated Stock Trading in Quantitative Finance, 2020 - Introduce a DRL library FinRL that facilitates beginners to expose themselves to quantitative finance and to develop their own stock trading strategies. Deep Reinforcement Learning for Automated Stock Trading: An Ensemble Strategy, 2020 - Propose an ensemble strategy that employs deep reinforcement schemes to learn a stock trading strategy by maximizing investment return. Courses & Books & Blogs ๐ QuantResearch - Quantitative analysis, strategies and backtests https://letianzj.github.io/ NYU: Overview of Advanced Methods of Reinforcement Learning in Finance Udacity: Artificial Intelligence for Trading AI in Finance - Learn Fintech Online. The Automation Ahead - CFA Institute examples for GenAI-driven investment workflows. Advanced-Deep-Trading - Experiments based on "Advances in financial machine learning" book. Advances in Financial Machine Learning - Using advanced ML solutions to overcome real-world investment problems. Build Financial Software with Generative AI - Book about how to build financial software hands-on using generative AI tools like ChatGPT and Copilot. Financial AI in Practice - A book about creating profitable, regulation-compliant financial applications. Investing for Programmers - A book about maximizing your portfolio, analyzing markets, and making data-driven investment decisions using Python and generative AI Mastering Python for Finance - Sources codes for: Mastering Python for Finance, Second Edition. MLSys-NYU-2022 - Slides, scripts and materials for the Machine Learning in Finance course at NYU Tandon, 2022. Train and Deploy a Serverless API to predict crypto prices - In this tutorial you won't build an ML system that will make you rich. But you will master the MLOps frameworks and tools you need to build ML systems that, together with tons of experimentation, can take you there. KeepRule - AI-powered investment discipline platform with principles from 26 legendary investors including Buffett, Munger, and Dalio. ai-agent-playbook - Knowledge base and runnable experiments on AI agent engineering, quantitative finance, and LLM-based factor research. Strategies & Research Time Series Data Price and Volume process with Technology Analysis Indices ๐๐ stockpredictionai - A complete process for predicting stock price movements. ๐ Personae - Implements and environment of Deep Reinforcement Learning & Supervised Learning for Quantitative Trading. ๐ Ensemble-Strategy - Deep Reinforcement Learning for Automated Stock Trading. FinRL - A Deep Reinforcement Learning Library for Automated Stock Trading in Quantitative Finance. AutomatedStockTrading-DeepQ-Learning - Build a Deep Q-learning reinforcement agent model as automated trading robot. tf_deep_rl_trader - Trading environment(OpenAI Gym) + PPO(TensorForce). trading-gym - Trading agent to train with episode of short term trading itself. trading-rl - Deep Reinforcement Learning for Financial Trading using Price Trailing. deep_rl_trader - Trading environment(OpenAI Gym) + DDQN (Keras-RL). Quantitative-Trading - Papers and code implementing Quantitative-Trading. gym-trading - Environment for reinforcement-learning algorithmic trading models. zenbrain - A framework for machine-learning bots. DeepLearningNotes - Machine learning in quant analysis. stock_market_reinforcement_learning - Stock market trading OpenAI Gym environment with Deep Reinforcement Learning using Keras. Chaos Genius - ML powered analytics engine for outlier/anomaly detection and root cause analysis.. mlforecast - Scalable machine learning based time series forecasting. patternity - Deterministic algorithm for stock price prediction, focusing on pattern recognition in historical data. Quantium Research - Research experiments exploring uncommon quant techniques. Portfolio Management Deep-Reinforcement-Stock-Trading - A light-weight deep reinforcement learning framework for portfolio management. qtrader - Reinforcement Learning for portfolio management. PGPortfolio - A Deep Reinforcement Learning framework for the financial portfolio management problem. DeepDow - Portfolio optimization with deep learning. ml-quant-trading - PyTorch research stack for mask-aware multi-factor modeling, ML baselines, portfolio optimization, and vectorized backtesting. skfolio - Python library for portfolio optimization built on top of scikit-learn. High Frequency Trading High-Frequency-Trading-Model-with-IB - A high-frequency trading model using Interactive Brokers API with pairs and mean-reversion. ๐ SGX-Full-OrderBook-Tick-Data-Trading-Strategy - Solutions for high-frequency trading (HFT) strategies using data science approaches (Machine Learning) on Full Orderbook Tick Data. HFT_Bitcoin - Analysis of High Frequency Trading on Bitcoin exchanges. Event Drive ๐๐ stockpredictionai - Complete process for predicting stock price movements. ๐ trump2cash - A stock trading bot powered by Trump tweets. Crypto Currencies Strategies LSTM-Crypto-Price-Prediction - Predicting price trends in crypto markets using an LSTM-RNN for trading. tforce_btc_trader - TensorForce Bitcoin trading bot. Tensorflow-NeuroEvolution-Trading-Bot - A population model that trade cyrpto and breed and mutate iteratively. gekkoga - Genetic algorithm for solving optimization of trading strategies using Gekko. Gekko_ANN_Strategies - ANN trading strategies for the Gekko trading bot. gekko-neuralnet - Neural network strategy for Gekko. bitcoin_prediction - Code for "Bitcoin Prediction" by Siraj Raval on YouTube. DeepAlpha - AI-powered crypto trading bot with 3-model ML ensemble (XGBoost, LightGBM, CatBoost), 12 exchanges via CCXT, walk-forward validated. Technical Analysis QTradeX - A powerful and flexible Python framework for designing, backtesting, optimizing, and deploying algotrading bots quant-trading - Python quantitative trading strategies. Gekko-Bot-Resources - Gekko bot resources. gekko_tools - Gekko strategies, tools etc. gekko RSI_WR - Gekko RSI_WR strategies. gekko HL - Calculate down peak and trade on. EthTradingAlgorithm - Ethereum trading algorithm using Python 3.5 and the library ZipLine. gekko_trading_stuff - Awesome crypto currency trading platform. forex.analytics - Node.js native library performing technical analysis over an OHLC dataset with use of genetic algorithmv. Bitcoin_MACD_Strategy - Bitcoin MACD crossover trading strategy backtest. crypto-signal - Automated crypto trading & technical analysis (TA) bot for Bittrex, Binance, GDAX, and more. Gekko-Strategies - Strategies to Gekko trading bot with backtests results and some useful tools. gekko-gannswing - Gann's Swing trade strategy for Gekko trade bot. Chartscout - Real-time cryptocurrency chart pattern detection with automated alerts using pattern recognition algorithms MarginSafe.ai - AI stock analysis platform specialized in intrinsic value and Wyckoff timing. Wickra - Native library computing 500+ technical-analysis indicators over OHLC data, streaming-first; Rust core with Python, Node.js, WASM and a C ABI (C/C++/C#/Go/Java/R) bindings. PineTS - Open-source Pine Script transpiler and runtime for indicators and strategies in Node.js and browsers, using live or custom market data. Lottery & Gamble LotteryPredict - Use LSTM to predict lottery. Arbitrage ArbitrageBot - Arbitrage bot that currently works on bittrex & poloniex. r2 - Automatic arbitrage trading system powered by Node.js + TypeScript. cryptocurrency-arbitrage - A crypto currency arbitrage opportunity calculator. Over 800 currencies and 50 markets. bitcoin-arbitrage - Bitcoin arbitrage opportunity detector. blackbird - Long / short market-neutral strategy. Data Sources Traditional Markets ๐ Quandl - Get millions of financial and economic dataset from hundreds of publishers via a single free API. yahoo-finance - Python module to get stock data from Yahoo! Finance. Tushare - TuShare is a utility for crawling historical data of China stocks. Congressional Stock Brain - Free AI-powered tool that scores U.S. STOCK Act congressional trade disclosures by significance. Committee weighting, timing analysis, 537 members tracked. Financial Data - Stock Market and Financial Data API. FinanceKit MCP - MCP server for stock quotes, technical analysis, crypto data, risk metrics, and portfolio analysis. No API keys for core data. Tapetide - Indian market data API and MCP server for NSE/BSE stocks, financials, screeners, and institutional flows. FXMacroData - Macroeconomic and FX data API with central bank announcements, policy rates, inflation, employment, GDP, release calendars, and MCP access for 18 currencies. StockAInsights - Institutional-grade financial statements API with AI extraction from SEC filings โ not XBRL. Covers domestic and foreign filers (20-F, 6-K, 40-F), normalized quarterly and annual data. 13F Insight - AI-powered 13F SEC filing tracker. Monitor hedge fund and institutional investor portfolio changes, with smart money move alerts and historical holding comparisons. ValueRay - Technical, quantitative and sentiment data for stocks and ETFs with risk metrics, peer percentiles and market regime signals. Optimized for AI/LLM agents. BenchGecko - AI economy tracking platform. Market cap, funding rounds, AI Bubble Index, company valuations, and compute supply chain data. FilingFirehose - SEC EDGAR JSON API with classified 8-Ks, activist 13D/G tagging, ATM offering detection, and hosted MCP access. SPZCO - Research terminal for U.S. congressional trades, executive disclosures, institutional holdings, and insider filings, with links to source documents. Edgrapi - Metered REST API and hosted MCP service for SEC filings, insider trades, institutional holdings, and company fundamentals. API key required. EarningsCall - API and SDKs for earnings call transcripts, audio, and presentation slides, with speaker identification and Q&A segmentation. The Stall - Pay-per-call API and MCP service aggregating stock, macroeconomic, crypto, and prediction-market data from third-party sources. Crypto Currencies BitBank.nz - AI-powered crypto forecasting and predictions API with machine learning models for 70+ cryptocurrency pairs. CryptoInscriber - A live crypto currency historical trade data blotter. Download live historical trade data from any crypto exchange. CoinPulse - Python SDK for cryptocurrency portfolio tracking with real-time prices, P/L calculations, backtesting, and price alerts. Free tier: 25 req/hr. Gekko-Datasets - Gekko trading bot dataset dumps. Download and use history files in SQLite format. Frostbyte Crypto API - Free real-time cryptocurrency price data API. Supports BTC, ETH, SOL, and 20+ tokens. No signup or API key required for basic endpoints. JSON responses with price, 24h change, market cap, and volume. CoinPaprika API - Cryptocurrency prices, volume, market cap, and historical data via API. Keyless access is available with limits; see plans . DexPaprika API - DEX pool, token, OHLCV, and trade data via API and MCP. Keyless access has rate and usage limits; see plans . Philidor - Institutional-grade DeFi risk scoring for 700+ vaults across 9 protocols and 6 chains. REST API and MCP server (Claude, Cursor, Windsurf). Deterministic 0โ10 risk scores, tiers (Prime/Core/Edge), portfolio analysis, oracle monitoring. No API key required. PreReason - Pre-analyzed financial market briefings optimized for AI agent consumption. 17 briefings covering BTC on-chain, macro (Fed balance sheet, M2, Treasury yields), and cross-asset correlations. Returns regime classification, trend signals, and confidence scores in markdown. Satoshi API - Bitcoin fee intelligence API with 108 endpoints for fee estimates, mempool analysis, block data, and mining stats. Self-hostable, Apache 2.0. Sharpe - Agent-ready crypto market intelligence API and MCP server for funding, derivatives, arbitrage, narratives, listings, and news. TBD Predict - Solana-based prediction market for human opinions with an agent CLI and AGENTS.md spec for AI agents to authenticate, list opinion campaigns, and place bets via JSON-friendly commands. 0xArchive - Real-time and historical Hyperliquid and Lighter market data via REST and WebSocket APIs. AgentServices - Crypto market data, on-chain analytics, and FX data via REST and MCP, with free endpoints and metered paid access. News Data WorldMonitor - AI-powered news aggregation, geopolitical monitoring, and infrastructure tracking in a unified situational awareness interface. Alternative Data Adanos Market Sentiment API - Market sentiment API for AI finance agents covering stocks across Reddit, X/Twitter, news, and Polymarket prediction markets with buzz, sentiment, trending, and comparison signals. Pizzint - Pentagon Pizza Index (PizzINT) is a real-time Pentagon pizza tracker that visualizes unusual activity at Pentagon-area pizzerias. It highlights a signal that has historically aligned with late-night, high-tempo operations and breaking news. Prediction Markets Parsec API - Unified prediction market infrastructure for normalized data, execution, and live streams across Polymarket, Kalshi, Opinion, Limitless, and PredictFun. MCP server for AI agent trading. Generous free tier. PolyMind - Real-time Polymarket trading alerts with multi-AI analysis (Groq, Claude, Gemini). Track whale bets, volume spikes, coordinated wallets, and 12 signal types. Free tier available. Live Tennis API - Tennis scores, win probabilities, and historical match results for prediction-market research via REST and WebSocket. History and advanced features require paid plans. Research Tools DDScore - AI-assisted first-pass due diligence for private-company materials, producing a structured 0โ100 score and written report across 12 dimensions while checking relevant claims against current public sources. Supports analyst judgement; not investment advice or a replacement for full due diligence. Synthical - AI-powered collaborative environment for Research. ๐๐ TensorTrade - Trade efficiently with reinforcement learning. ML-Quant - Quant resources from ArXiv (sanity), SSRN, RePec, Journals, Podcasts, Videos, and Blogs. JAQS - An open source quant strategies research platform. pyfolio - Portfolio and risk analytics in Python. alphalens - Performance analysis of predictive (alpha) stock factors. empyrical - Common financial risk and performance metrics. Used by Zipline and pyfolio. zvt - Zero vector trader. CongressionalStockBrain - AI-powered tool that ingests U.S. STOCK Act congressional trade disclosures and converts them into machine-scored signals for retail investors. WalletLens - Multi-asset portfolio tracker with AI insights, technical analysis, live prices, and local-first data storage. FN2 - AI market analyst for investors: personalized daily briefings, deep stock research, and scheduled agents that watch your watchlist and earnings. Free tier available. WFGY โ Open source framework for debugging and stress testing LLM agents and RAG pipelines. Includes a 16 mode failure map and long-horizon stress tests that are useful for financial research agents. ChainPulse - AI-powered Bitcoin network intelligence CLI for natural language queries on mempool, fees, blocks, and mining analysis. CRNG - Contingency RNG, generates random numbers with real market fat tails (K=5-220) and volatility clustering. Matches 86% of real market metrics vs 14% for NumPy. Includes regime detector. Chart Library - Visual chart pattern search engine. Upload a screenshot or type a ticker+date to find the 10 most similar historical chart patterns and see what happened next. 24M+ embeddings, 19K symbols, REST API + MCP server. Coinugget - Real-time RSI signals, price action & volume spikes dashboard for crypto traders. Free, no sign-up required. QuantLink - AI-powered US-equity research terminal with deep-research agents grounded in SEC filings, a fundamental and technical stock screener with backtesting, institutional 13F holder analysis, insider (Form 4) activity, and congressional trade tracking from STOCK Act disclosures. Free tier available. NeuPortal - AI forecasting-accountability lab: every forecast is locked pre-event, Bitcoin-timestamped (OpenTimestamps), and Brier-scored against prediction markets in public. Trading System For Back Test & Live trading Traditional Market System the0 - Self-hosted execution engine for algorithmic trading bots. Supports Python, TypeScript, Rust, C++, C#, Scala, and Haskell. Each bot runs in an isolated container with scheduled or streaming execution. ๐๐๐ OpenBB - AI-powered opensource research and analytics workspace. ๐๐ zipline - A python algorithmic trading library. ๐ TradingView - Get real-time information and market insights. rqalpha - A extendable, replaceable Python algorithmic backtest & trading framework. backtrader - Python backtesting library for trading strategies. kungfu - Kungfu Master trading system. finclaw - AI-native quantitative trading engine with 484 alpha factors, genetic algorithm strategy evolution, walk-forward backtesting and paper trading. Supports A-shares, crypto, and MCP server for AI agent integration. lean - Algorithmic trading engine built for easy strategy research, backtesting and live trading. Combine & Rebuild pylivetrader - Python live trade execution library with zipline interface. CoinMarketCapBacktesting - As backtest frameworks for coin trading strategy. Crypto Currencies zenbot - Command-line crypto currency trading bot using Node.js and MongoDB. bot18 - High-frequency crypto currency trading bot developed by Zenbot. magic8bot - Crypto currency trading bot using Node.js and MongoDB. catalyst - An algorithmic trading library for Crypto-Assets in python. QuantResearchDev - Quant Research dev & Traders open source project. MACD - Zenbot MACD Auto-Trader. abu - A quant trading system base on python. Plugins CoinMarketCapBacktesting - Tests bt and Quantopian Zipline as backtesting frameworks for coin trading strategy. Gekko-BacktestTool - Batch backtest, import and strategy params optimalization for Gekko Trading Bot. TA Lib pandas_talib - A Python Pandas implementation of technical analysis indicators. finta - Common financial technical indicators implemented in Python-Pandas (70+ indicators). tulipnode - Official Node.js wrapper for Tulip Indicators. Provides over 100 technical analysis overlay and indicator functions. techan.js - A visual, technical analysis and charting (Candlestick, OHLC, indicators) library built on D3. Exchange API Do it in real world! Trade It - MCP for trading on common brokerages (Robinhood, ETrade, Schwab, Webull, Public, tastytrade, Coinbase, Kraken so far) IbPy - Python API for the Interactive Brokers on-line trading system. HuobiFeeder - Connect HUOBIPRO exchange, get market/historical data for ABAT trading platform backtest analysis and live trading. ctpwrapper - Shanghai future exchange CTP api. PENDAX - Javascript SDK for Trading/Data API and Websockets for cryptocurrency exchanges like FTX, FTXUS, OKX, Bybit, & More Framework tf-quant-finance - High-performance TensorFlow library for quantitative finance. Visualizing playground - Play with neural networks. netron - Visualizer for deep learning and machine learning models. KLineChart - Highly customizable professional lightweight financial charts GYM Environment ๐ TradingGym - Trading and Backtesting environment for training reinforcement learning agent. TradzQAI - Trading environment for RL agents, backtesting and training. btgym - Scalable, event-driven, deep-learning-friendly backtesting library. TraderHarness - Contamination-resistant A-share backtesting environment for LLM trading agents, with point-in-time masking, entity/date anonymization, fingerprinted replay, and trajectory (SFT) export. Articles The-Economist - The Economist. nyu-mlif-notes - NYU machine learning in finance notes. Using LSTMs to Turn Feelings Into Trades Others zipline-tensorboard - TensorBoard as a Zipline dashboard. gekko-quasar-ui - An UI port for gekko trading bot using Quasar framework. Floom AI gateway and marketplace for developers, enables streamlined integration and least volatile approach of AI features into products LendTrain - AI-native mortgage refinance plugin for Claude Code with real-time institutional pricing, state-specific closing costs, FHA Streamline/VA IRRRL detection, and regulatory compliance. Uses MCP (Model Context Protocol) to connect LLMs to live mortgage pricing. Registry Broker - Universal AI agent index for discovering trading agents across Virtuals Protocol, NANDA, MCP, and other registries. KeepRule - AI-powered investment discipline tracking platform with curated principles from 26 legendary investors including Buffett, Munger, and Dalio. Helps traders maintain rational decision-making. Philidor - DeFi risk infrastructure for AI agents: MCP server and REST API for vault risk scores, portfolio analysis, and due diligence. No API key. 700+ vaults, 9 protocols, 6 chains. Hindsight - State-of-the-art long-term memory for AI agents by Vectorize. Open source, self-hostable, with integrations for LangChain, CrewAI, MCP, and more. Gives financial trading agents persistent memory across sessions. Other Resource ๐๐๐ Stock-Prediction-Models - Stock-Prediction-Models, Gathers machine learning and deep learning models for Stock forecasting, included trading bots and simulations. ๐๐ Financial Machine Learning - A curated list of practical financial machine learning (FinML) tools and applications. This collection is primarily in Python. ๐ Awesome-Quant-Machine-Learning-Trading - Quant / Algorithm trading resources with an emphasis on Machine Learning. awesome-quant - A curated list of insanely awesome libraries, packages and resources for Quants (Quantitative Finance). FinancePy - A Python Finance Library that focuses on the pricing and risk-management of Financial Derivatives, including fixed-income, equity, FX and credit derivatives. Explore Finance Service Libraries & Projects - Explore a curated list of Fintech popular & new libraries, top authors, trending project kits, discussions, tutorials & learning resources on kandi. AgentMarket - B2A marketplace for AI agents. 189 listings, 28M+ real energy data records, LangChain/MCP integration. MeterCall โ One metered API gateway. 21M+ endpoints (payments, SMS, AI, CRMs, gov data). Free tier; pay per call.
2. Related Work, Literature Survey & Algorithmic Foundations
Awesome AI for Finance A curated list of awesome AI / machine learning / deep learning / reinforcement learning / LLM tools, frameworks, and resources for quantitative finance. This list is maintained by the AI4Finance Foundation . If you find it useful, please give it a star๐! Contents AI4Finance Foundation Ecosystem LLMs for Finance AI Agents for Finance Deep Reinforcement Learning Machine Learning Quantitative Trading & Backtesting Data Sources Technical Analysis Portfolio Management & Risk Sentiment Analysis & NLP Quantitative Finance Libraries Trading Platforms Visualization & Rendering High Performance Computing Research Papers & Surveys Books Contributing AI4Finance Foundation Ecosystem The AI4Finance Foundation provides a full-stack open-source ecosystem for AI-driven finance, from data to deployment. Project Stars Description FinGPT 20.4k Open-source financial large language models. Fine-tuning LLMs (LLaMA, ChatGLM, etc.) with financial data for sentiment analysis, robo-advising, and quantitative trading. FinRL 15.3k The first open-source framework for financial reinforcement learning. Full pipeline from data processing to DRL-based trading (stock, crypto, forex). FinRobot 7.2k Open-source AI agent platform for financial analysis using LLMs. Multi-agent architecture for report generation, analysis, and trading decisions. ElegantRL 4.3k Scalable and elastic deep reinforcement learning library using PyTorch. Lightweight, efficient, and cloud-native. FinRL-X 3.2k AI-native modular infrastructure for quantitative trading. Unifies data, strategy, backtesting, and execution. Paper FinRL-Meta 1.9k A universe of dynamic market environments for financial reinforcement learning. Standardized datasets for stock, crypto, and forex. FinNLP 1.5k Democratizing internet-scale financial data. NLP toolkit for financial news and social media data processing. FinRL-Tutorials 1.2k Jupyter notebook tutorials covering stock trading, portfolio allocation, crypto trading, and more. FinRL-Podracer 501 Cloud-based framework for massively parallel financial RL training. RLSolver 167 RL-based solver for combinatorial optimization problems in finance. Helium MCP โ Real-time news with 37-dimension bias scoring, ML options pricing, and live market data. Interactive demo ยท REST API LLMs for Finance Large Language Models are transforming finance through automated analysis, report generation, and intelligent trading. Project Stars Description FinGPT 20.4k Open-source financial LLM framework with LoRA/QLoRA fine-tuning on financial data. FinRobot 7.2k AI agent platform for financial analysis using LLMs with multi-agent architecture. FinGLM 2.2k Open financial LLM by Zhipu AI built on ChatGLM, with annual report datasets and challenge competition code. FinBERT 2.1k Pre-trained NLP model to analyze sentiment of financial text. Built on BERT, fine-tuned on financial corpus. FinNLP 1.5k Toolkit for financial NLP with easy access to financial news and social media data. XuanYuan 1.3k Chinese financial chat LLM from Du Xiaoman Financial. Hundreds of billions of parameters, fine-tuned on financial domain data. DISC-FinLLM 878 Chinese financial LLM from Fudan University. Multi-expert fine-tuning for QA, domain NLP, math computation, and RAG-based consulting. PIXIU (FinMA) 865 Open-source financial LLM evaluation framework with FinMA models (7B/30B), FinBen benchmark, and 136K instruction-tuning dataset. Fin-R1 790 Financial reasoning LLM based on Qwen2.5-7B, trained with SFT + RL on 60K chain-of-thought samples. SOTA on FinQA and ConvFinQA. InvestLM 153 Investment-focused LLM fine-tuned on LLaMA-65B using CFA exams, SEC filings, and quant finance data. Rated comparable to GPT-4 by hedge fund managers. AI Agents for Finance | Pineify | โ | AI-assisted trading workflow with Pine Script, MQL5, and cTrader coding agents, financial research tools, market screening, strategy optimization, and backtest analysis. | Autonomous AI agents that can analyze markets, make trading decisions, and generate financial reports. TradingAgents vs TradingAgents-CN vs Vibe-Trading - Side-by-side comparison of open-source LLM multi-agent trading frameworks. Project Stars Description iPulse AI โ Open Agentic Investment Research Platform with independent AI advisor reports, asset forecasts and ranked market discovery; evidence, risks and assumptions are inspectable. Hosted research software, with no trade execution. TradingAgents 82.4k Multi-agent LLM financial trading framework. Specialized agents (fundamentals, sentiment, technical) debate and reach consensus on trades. ai-hedge-fund 59.7k Multi-agent AI hedge fund simulator with 18 agents modeled after legendary investors (Buffett, Munger, etc.) plus specialist agents. Lumina AI 6 Autonomous multi-agent AI hedge fund combining Qlib Alpha158 quantitative ML, LangGraph consensus deliberation, and a deterministic 10-second fast watchdog sentinel. daily_stock_analysis 40.0k LLM-powered A/H/US stock analysis system. Multi-source market data + real-time news + LLM decision dashboard + multi-channel push notifications. Anthropic Financial Services 29.6k Anthropic's official reference agents for finance: 10 specialized agents for earnings review, KYC screening, pitch generation, model building, and more. TradingAgents-CN 27.8k Chinese multi-agent LLM trading framework simulating professional trading firm workflows for A/HK/US markets. AI-Trader 19.1k Fully-automated agent-native trading platform from HKU. Cross-platform signal sync and one-click copy trading. Qbot 17.5k AI-powered quantitative investment research platform with 300+ models from 40+ papers, fully local deployment. nofx 12.5k AI trading terminal for stocks, commodities, forex, and crypto. 7 exchanges + 7 AI models competing in real-time. Vibe-Trading 9.6k Personal trading agent from HKU for research, simulation, backtesting, and autonomous broker-authorized trading. FinRobot 7.2k Open-source AI agent platform for financial applications powered by LLMs. Polymarket Agents 3.6k Official Polymarket framework for building autonomous AI agents that trade prediction markets. AutoHedge 3.1k Swarm-intelligence-powered autonomous hedge fund with Director, Quant, Risk, and Execution agents. AlphaGen 1.1k Generates synergistic formulaic alpha factors via reinforcement learning (KDD 2023). QuantaAlpha 1.0k LLM + evolutionary strategies framework for automatic alpha factor mining. FinMem 903 LLM-based trading agent with layered memory (short/mid/long-term) for evolving market understanding. oracle3 258 Open-source trading engine and MCP server for prediction markets (Kalshi, Polymarket, Solana). Checks fee-adjusted no-arbitrage relations across related event contracts, runs statistical-arbitrage strategies, and trades live or on paper; the MCP server lets AI agents read order books and place paper orders. QVeris 230 Capability routing network for AI agents to discover, inspect, and call 10,000+ real-world financial capabilities through one unified MCP protocol. Superhighway M&A Research Agent โ Python agent that researches comparable M&A transactions, deal multiples, potential acquirers, and recent deal activity via live web search. Generates structured M&A intelligence briefs for corporate development, PE analysts, and investment bankers. Pay-per-call, no signup. NeuPortal โ AI forecasting-accountability lab. Every AI forecast is locked pre-event, Bitcoin-timestamped via OpenTimestamps, and Brier-scored in public against prediction markets across crypto, macro, and sports. Hosted, closed-source service. | Algorier | โ | Natural-language ("vibe trading") agent that turns a described trading idea into a generated, backtested and forward-tested algorithm and deploys it to the user's own broker account across forex, crypto, metals, indices, CFDs and equities; strategies can also be sold on its marketplace with the logic kept private from buyers. | Deep Reinforcement Learning Frameworks and libraries applying deep reinforcement learning to trading, portfolio management, and market making. Project Stars Description FinRL 15.3k The first open-source DRL framework for quantitative finance. Supports PPO, A2C, DDPG, SAC, TD3, and more. TensorTrade 6.3k An RL framework for training, evaluating, and deploying robust trading agents. ElegantRL 4.3k Massively parallel deep reinforcement learning using PyTorch. FinRL-Trading 3.2k Ensemble strategy using DRL for automated stock trading (ICAIF 2020). gym-anytrading 2.4k OpenAI Gym trading environment for reinforcement learning. FinRL-Meta 1.9k Dynamic market environments and standardized datasets for financial RL. Machine Learning General machine learning approaches for financial prediction, alpha generation, and quantitative investment. Project Stars Description Qlib 44.0k Microsoft's AI-oriented quantitative investment platform with full ML pipeline. Supports alpha mining, model training, and backtesting. ML for Trading 18.7k Code for the book Machine Learning for Algorithmic Trading . Comprehensive guide on using ML to add value to trading strategies. Stock-Prediction-Models 9.4k Collection of ML and deep learning models for stock forecasting, including transformers, LSTMs, and more. ai_quant_trade 5.7k Comprehensive AI stock trading platform covering LLMs, factor mining, ML/DL/RL, graph networks, and HFT strategies. stockpredictionai 5.6k Complete process for predicting stock price movements using deep learning. TF Quant Finance 5.4k TensorFlow-based library for quantitative finance by Google. Covers options pricing, volatility models, and more. QuantsPlaybook 5.2k Reproduces 100+ quantitative investment strategies from Chinese brokerage research reports. MLFinLab 4.8k Implementations of methods from Advances in Financial Machine Learning by Marcos Lopez De Prado. Adv_Fin_ML_Exercises 1.9k Experimental solutions to exercises from Advances in Financial Machine Learning . AlphaPy 1.7k ML framework for both speculators and data scientists. Automated feature engineering and model selection. fin-ml 1.2k Code for case studies in Machine Learning and Data Science Blueprints for Finance . ml-quant-trading 38 Paper-backed PyTorch research stack for mask-aware multi-factor modeling, ML baselines, portfolio optimization, and vectorized backtesting. predixlab-signals Daily US stock signals from LightGBM models (Pivot turning-point events and AlphaRank cross-sectional rankings) with expanding-window walk-forward backtests. Open source (Apache-2.0); market data is not included. Quantitative Trading & Backtesting Frameworks for algorithmic trading, strategy backtesting, and automated trading bots. Project Stars Description FreqTrade 51.1k Free, open-source crypto trading bot. Supports backtesting, strategy optimization, and live trading on many exchanges. vnpy/VeighNa 41.3k Python-based open-source quant trading platform. Multi-market gateway support for stocks, futures, options, and crypto. NautilusTrader 23.3k Production-grade Rust-native trading engine with nanosecond-resolution clock and Python strategy API. Backtrader 21.8k Feature-rich Python framework for backtesting and live trading. Supports multiple data feeds and brokers. Zipline 19.8k Pythonic event-driven backtesting library, originally powering the Quantopian platform. QuantConnect/Lean 19.7k Algorithmic trading engine built for strategy research, backtesting, and live trading. Multi-asset support. Hummingbot 18.8k Open-source crypto market making and arbitrage bot. Gateway to CeFi and DeFi exchanges. Abu 17.4k Python quant trading system supporting US/A/HK stocks, futures, options, and Bitcoin with ML-based strategy optimization. QUANTAXIS 10.6k Distributed quant solution for stocks/futures/options with data, backtesting, simulation, and live trading. Easytrader 9.8k Python stock trading automation for Chinese brokers (Tonghuashun/MiniQMT/Xueqiu). Jesse 8.0k Advanced algo-trading framework for crypto. Strategy development, backtesting, and live trading. VectorBT 7.8k Blazing fast vectorized backtesting and trading analysis using NumPy and Pandas. ManifoldBT High-performance Rust-powered Python backtesting: vectorized signals with realistic fills (fees, slippage, look-ahead), parameter sweeps, walk-forward, Monte Carlo. RQAlpha 6.4k Extendable Python algorithmic backtesting and trading framework from RiceQuant. WonderTrader 6.1k High-performance C++ core quant framework with Python layer (wtpy) for research, trading, and operations. OctoBot 6.0k Open-source crypto trading bot for AI, Grid, DCA, and TradingView strategies on 15+ exchanges. Krypto-trading-bot 3.7k Self-hosted C++ high-frequency crypto market-making bot with sub-millisecond order execution. Hikyuu 3.2k C++/Python ultra-high-speed quant research framework for A-share market. bt 2.9k Flexible backtesting framework for Python with modular algorithm stacks and portfolio-level strategy testing. HFT-LOB-Trading-ML 2.3k High-frequency trading strategies using ML on full order book tick data. barter-rs 2.2k Open-source Rust framework for building event-driven live-trading and backtesting systems. Zipline-Reloaded 1.8k Community-maintained fork of Zipline, compatible with modern Python. Companion to ML for Algorithmic Trading . Lumibot 1.6k Backtestable AI trading agents for stocks, options, crypto, futures, and forex with real broker integration. AutoTrader 1.3k Python-based platform for development, optimization, and deployment of automated trading systems. Funcat 1.0k Ports Tongdaxin/Tonghuashun financial formula syntax to Python for technical analysis (e.g., MA(C,5) ). Data Sources Tools and APIs for accessing financial market data โ stocks, crypto, forex, and alternative data. Project Stars Description OpenBB Terminal 68.5k Open-source investment research platform. Access stock, crypto, forex, and macro data with built-in analysis tools. CCXT 42.8k Unified JavaScript/Python/PHP library for 100+ cryptocurrency exchange APIs. Trading and market data. FinceptTerminal 25.1k Modern finance terminal with advanced market analytics, investment research, and economic data tools for interactive exploration. yfinance 24.0k Download market data from Yahoo Finance. Stocks, ETFs, mutual funds, options, and more. AKShare 20.0k Elegant Python financial data interface covering A-shares, futures, options, funds, forex, bonds, indices, and crypto from authoritative Chinese sources. TuShare 15.1k Financial data interface for Chinese stock markets. Historical and real-time data. OpenStock 13.0k Free open-source market platform alternative with real-time prices, alerts, and company insights. StockSharp 10.0k Algorithmic trading and quantitative analysis platform. Connects to multiple brokers and exchanges. FinanceDatabase 7.8k Comprehensive database of 300,000+ symbols: Equities, ETFs, Funds, Indices, Currencies, and Crypto. FinanceToolkit 4.9k Transparent financial analysis with 150+ ratios, indicators, and performance measurements. AData 4.6k Free open-source A-share quant trading database with multi-source data fusion and dynamic proxy. Binance 2.9k Official Python connector for Binance APIs. Spot, futures, and market data. ArcticDB 2.4k High-performance serverless DataFrame database by Man Group for petabyte-scale time-series and tick data. EdgarTools 2.3k Read and analyze SEC EDGAR filings in Python (10-K, 8-K, XBRL, 13F). FinRL-Meta 1.9k Dynamic datasets and market environments for stock, crypto, and forex. Alpaca 1.4k Official Python SDK for Alpaca. Commission-free stock trading API with paper/live trading. jqdatasdk 1.3k JoinQuant's Python data SDK for Chinese financial market data. DeFiLlama Adapters 1.2k Open-source DeFi TVL calculation adapters for 2,400+ protocols across 180+ blockchains. WRDS 155 Python interface to Wharton Research Data Services for academic financial research. roicai-python 0 Python SDK for roic.ai: financial statements, 90+ ratios and earnings call transcripts for 68k+ companies, as pandas or polars DataFrames. 0xArchive โ Real-time and historical Hyperliquid and Lighter market data via REST and WebSocket APIs. The Stall MCP server with 173 financial market data capabilities โ stocks, ETFs, crypto, DeFi, options, SEC filings, earnings, analyst ratings, and macro indicators. Pay-per-call via x402 micropayments on Base; no API keys required. AgentServices 54-service x402-paid API platform for AI agents โ crypto market data, onchain analytics, financial intelligence, and inference endpoints. MCP server with 37 tools. Pay-per-call via x402 on Base; no API keys required. S&P 500 Earnings Announcement Times 63,969 S&P 500 earnings announcements with the time of day, parsed from SEC 8-K item 2.02 filings, 2003-2026. CSV dataset released under CC0. Eulerpool โ Financial data API with 400+ endpoints: global equities and fundamentals, ETFs, insider trades, 13F, macro series from FRED/ECB/IMF/World Bank/OECD/BIS, crypto, FX and commodities. SDKs for Python, TypeScript, Go, Java, R, PHP, Rust and C++; hosted MCP server. Free tier 100k requests/month. aikstockdata Korean listed-company data as static JSON and CSV with a remote MCP server: the previous trading day's settled closing prices from the government equity price feed (ETFs excluded), DART regulatory filings with the minute each one was received, earnings from periodic filings, and median market-adjusted returns after a filing grouped by filing type. No account and no API key โ Non-commercial use with attribution; commercial redistribution is not permitted. FXMacroData โ Official-source macroeconomic releases for 22 currencies, each stamped with its publication time, plus release calendars, central-bank decisions, COT positioning and FX rates over REST and a hosted MCP server. USD data needs no API key. Equibles โ US stock market data over a hosted MCP server and a REST API: SEC filings with full-text search, XBRL financial statements, earnings call transcripts, insider and congressional trades, 13F holdings and option chains. Works with ChatGPT and Claude over OAuth. Free tier 100 requests/day. Aperiodic โ Point-in-time crypto microstructure, liquidity and order-flow metrics (220 metrics, 19 datasets) and raw trades, quotes and derivatives data for perpetuals on venues including Binance, OKX and Hyperliquid, as parquet via CLI, REST API and Python SDK. A preview of every dataset needs no API key. Fincept MCP โ Hosted MCP server for Fincept Terminal over Streamable HTTP with OAuth sign-in: quotes, candles, option chains, futures curves, fundamentals, economic series, SEC filings, news, backtests, paper trading and 15 quant engines (statistics, forecasting, GARCH volatility, portfolio optimisation, derivatives pricing). Requires a paid plan. Tapetide โ Indian stock market data for about 8,200 NSE and BSE listed companies over a hosted MCP server: quotes, quarterly financials, shareholding, a 326-ratio screener, FII/DII flows, option chains and company filings. Works with Claude, ChatGPT and Cursor over OAuth or a personal token; open-source MIT client on GitHub . Free plan included. Financial Evidence Read-only MCP and Python tools for source-linked institution, funding and market-liquidity research, with persistent clients and source-status SSE . Retains source dates, unavailable states and monitor expiry. Public reads need no API key; the stream is metadata-only and does not execute trades. Invokely โ Hosted MCP and REST API for official-source macroeconomic indicators such as GDP, inflation, employment, and interest rates. Each value includes its reporting period and publisher citation; coverage includes U.S. states and counties, EU regions, Spanish municipalities, and World Bank country data. Edgrapi โ US government and SEC data as JSON over REST and a hosted MCP server: SEC EDGAR insider Form 4 trades, 8-K events, 13F holdings, 13D/G stakes and XBRL fundamentals, House STOCK Act congressional trades, plus SAM.gov contract opportunities, USAspending awards and Grants.gov grants. OAuth or API key; 50 free credits to start, no card. Ryufin โ Free US stock research from SEC EDGAR filings for 3,000+ listed companies, as a website and a hosted MCP server (Streamable HTTP at https://ryufin.com/mcp ): five plain answers per stock (business quality, safety, valuation, growth, market) with the checks and filing dates behind them, reported financial series, 8-K events and the answers that changed in the latest nightly run. No account and no API key; 60 requests a minute per address. EDGAR Supply Chain Major customers that 3,049 US-listed companies disclose in their latest 10-K, 20-F or 40-F (the ASC 280 10% customer rule), with the share of revenue and a link to the filing, plus a 322-edge supplier-customer graph between listed companies. CSV and JSON under CC BY 4.0; MIT-licensed Python extractor over XBRL tags, tables and filing text. Level Open Finance Read-only MCP server and 20 inspectable SMB finance and month-end close skills ( SKILL.md + examples/index.json synthetic fixtures) covering bank-to-ledger reconciliation, 13-week cash-flow forecasting, AR integrity, vendor 1099 readiness, and CFO variance review. MIT-licensed; guided walkthroughs at levelcfo.com/finance-systems . LevelBench-100 & SMB Financial Benchmarks 100-check AI accounting and CFO benchmark evaluating 38 AI models and CLI harnesses across 15 anonymized Excel ( .xlsx ) close workbooks, 31 task rubrics, and 30 AI accounting error modes ( levelcfo.com/benchmarks/agentic-accounting ), alongside empirical SMB financial percentiles across 2,235 US service businesses and 6 SEC 10-K public contractor comps ( HF Datasets ). Technical Analysis Libraries for computing technical indicators and performing quantitative analysis on market data. Project Stars Description TA-Lib 12.0k Python wrapper for TA-Lib. 200+ technical analysis functions: candlestick patterns, momentum, volatility, and more. ta 5.1k Technical analysis library using Pandas and NumPy. Includes 40+ indicators out of the box. Clairvoyant 2.4k Identify and monitor social/historical cues for short-term stock movement. Funcat 1.0k Tongdaxin/Tonghuashun financial formula syntax ported to Python (e.g., MA(C,5) , CROSS(MA5, MA10) ). Portfolio Management & Risk My AI Investment OS - Free 5-question diagnostic that identifies your investing school (Index/Value/Growth/Trend), delivers an 8-AI committee verdict, and generates a personalized Prompt Library. Includes 6 companion tools: DCA Simulator, Recovery Navigator, Dividend Engine, Kelly Master, Pyramid Builder, Portfolio Clarity. No login required. Tools for portfolio optimization, risk analysis, asset allocation, and performance analytics. Project Stars Description QuantStats 7.2k Portfolio analytics for quants with in-depth performance metrics, risk analytics, tear sheets, and HTML reports. Pyfolio 6.3k Portfolio and risk analytics in Python with comprehensive tear sheet analysis. PyPortfolioOpt 5.8k Portfolio optimization in Python. Classical efficient frontier, Black-Litterman, Hierarchical Risk Parity, and more. Riskfolio-Lib 4.2k Portfolio optimization and quantitative strategic asset allocation. Supports 20+ risk measures. ffn 2.6k Financial functions library for Python: performance measurement, data transformations, and portfolio analytics. Empyrical 1.5k Common financial risk and performance metrics (Sharpe, Sortino, max drawdown). Used by Zipline and Pyfolio. OLPS 357 Toolbox for On-Line Portfolio Selection strategies. Sentiment Analysis & NLP Tools for extracting trading signals from text โ news, social media, SEC filings, and earnings calls. Project Stars Description FinGPT 20.4k Financial LLMs with built-in sentiment analysis capabilities via fine-tuning. VADER 5.0k Valence Aware Dictionary and sEntiment Reasoner. Rule-based sentiment analysis tool, widely used for financial social media and news text. FinBERT 2.1k Pre-trained NLP model for financial sentiment analysis. Fine-tuned on financial communication text. Quantitative Finance Libraries Core quantitative finance libraries for derivatives pricing, risk management, and financial modeling. Project Stars Description gs-quant 10.5k Goldman Sachs Python toolkit for quantitative finance, derivatives pricing, and risk analytics. QuantLib 7.2k The canonical open-source C++ library for quantitative finance: derivatives pricing, fixed income, and risk management. RustQuant 1.8k Rust library for quantitative finance covering stochastic processes, options pricing, bonds, and Monte Carlo methods. Trading Platforms End-to-end platforms for algorithmic trading with broker integration. Project Stars Description QuantConnect/Lean 19.7k Algorithmic trading engine. Multi-asset support with cloud integration for backtesting and live trading. StockSharp 10.0k Algorithmic trading for stocks, forex, crypto, and options. GUI and API-based development. FinRL-X 3.2k AI-native modular infrastructure for quantitative trading. Unifies data processing, strategy construction, backtesting, and broker execution with RL/LLM support. Paper Visualization & Rendering Tools for visualizing market data, trading activity, and agent performance. Project Stars Description KLinePic MCP First-party MCP server and Agent API for turning broker or exchange fills into annotated candlestick trade-review charts. Supports A-shares, US stocks, crypto, and futures; post-trade review only, with no signals or order execution. mplfinance 4.4k Financial data visualization using Matplotlib. Candlestick charts, OHLC, Renko, and more. TradingGym 1.9k Toolkit for training and backtesting RL algorithms with real-time rendering. Kandle Browser charting app for crypto, stocks and forex with 36 indicators, scripts with strategy backtests, alerts and bar replay. Its MCP server (26 tools) lets AI agents read candles and indicator values, draw on the chart, run scripts and take screenshots. No account; crypto data needs no API key. High Performance Computing Project Stars Description NumPy 32.1k The fundamental package for scientific computing with Python. Foundation for most quantitative finance libraries. Azure HPC - Azure high-performance computing for financial services โ risk simulation, pricing, and analytics. Research Papers & Surveys Key academic papers on AI in finance. Surveys Hambly, Ben, Renyuan Xu, and Huining Yang. " Recent Advances in Reinforcement Learning in Finance ." arXiv preprint arXiv:2112.04553 , 2021. " A Survey of Large Language Models for Financial Applications ." arXiv , 2024. Covers FinGPT, BloombergGPT, and domain-specific fine-tuning. " FinAgent: A Multimodal Foundation Agent for Financial Trading ." arXiv , 2024. Multimodal agent processing text, tables, and charts. " FinRobot: An Open-Source AI Agent Platform for Financial Applications ." arXiv , 2024. " Data-Centric FinGPT: Open-Source Financial Large Language Models ." arXiv , 2023. Data-centric approach to building financial LLMs. Selected Papers " Deep Reinforcement Learning for Automated Stock Trading: An Ensemble Strategy ." ICAIF 2020 . The ensemble strategy behind FinRL-Trading. " FinRL: A Deep Reinforcement Learning Library for Automated Stock Trading in Quantitative Finance ." NeurIPS 2020 Deep RL Workshop . " FinRL-Meta: Market Environments and Benchmarks for Data-Driven Financial Reinforcement Learning ." NeurIPS 2022 . " FinMem: A Performance-Enhanced LLM Trading Agent with Layered Memory ." arXiv , 2023. Novel memory architecture for LLM trading agents. " FinGPT: Open-Source Financial Large Language Models ." arXiv , 2023. Democratizing financial LLMs with a data-centric approach and accessible fine-tuning. " TradingAgents: Multi-Agents LLM Financial Trading Framework ." arXiv , 2024. " CAST: A Cross-Asset State-Space Trading System for Drawdown Control in Stock Markets ." ICDM 2026 . Cross-asset Kalman filtering with model predictive control for drawdown-aware trading. Code " AgonAlpha: Autonomous Alpha Discovery via Prompt Economy and Scalable Agentic Search ." NeurIPS 2026 Workshop (Agenthon) . The first LLM alpha-mining system to combine verified research-artifact search, a fresh-context adversarial reviewer with veto power, and parallel budget allocation; SPECTACULAR-grade alphas on WorldQuant BRAIN with Sharpe up to 3.48. Code " LLMs Meet Finance: Fine-Tuning Foundation Models for the Open FinLLM Leaderboard ." arXiv , 2025. SFT, DPO and synthetic-data RL on Qwen2.5 and DeepSeek-R1 across financial tasks; DPO cuts overlong outputs from 54.7% to 1.7%. Books Giller, Graham L. Adventures in Financial Data Science . Giller Investments, 2020. Jansen, Stefan. Machine Learning for Algorithmic Trading . 2nd ed. Packt, 2020. Code De Prado, Marcos Lopez. Advances in Financial Machine Learning . Wiley, 2018. Tatsat, Hariom, Sahil Puri, and Brad Lookabaugh. Machine Learning and Data Science Blueprints for Finance . O'Reilly, 2020. Code Dixon, Matthew F., Igor Halperin, and Paul Bilokon. Machine Learning in Finance: From Theory to Practice . Springer, 2020. Contributing Contributions are welcome! If you have suggestions for new projects or improvements: Fork this repository. Add your project in the appropriate section, following the existing format. Submit a pull request with a brief description of the addition. Selection criteria : Projects should be actively maintained, well-documented, and provide genuine value to the AI for finance community. We particularly welcome: Open-source tools with demonstrated community adoption Research projects with accompanying papers Frameworks that solve real problems in quantitative finance License This project is licensed under the MIT License . Feedback : If you have any ideas or want content added to this list, feel free to open an issue or submit a PR . BDE Scoreโข - AI-Powered Multi-Market Stock Analysis โ transparent multi-factor scoring for 73 stocks across US, HK, and A-share markets. EU AI Act Art.50 compliant.
3. Dataset Ingestion, Preprocessing & Stochastic Augmentations
Machine Learning for Finance A CQF elective with Dr. Yves J. Hilpisch, The Python Quants GmbH General resources: http://tpq.io http://hilpisch.com Wifi You can use the following wifi in this venue: SSID Fitch Guest PW Ft1ch#2016#! Abstract This CQF elective is about machine learning and deep learning with Python applied to finance. It starts with techniques to retrieve financial data from open data sources and covers Python packages like NumPy, pandas, scikit-learn and TensorFlow. It provides the basis to further explore these recent developments in data science to improve traditional financial tasks such as the pricing of American options or the prediction of future stock market movements. Among other, the elective covers: Getting and working with financial time series data in Python Using linear OLS regression to predict financial prices & returns Using scikit-learn for machine learning with Python Application to the pricing of American options by Monte Carlo simulation Applying logistic regression to classification problems Predicting stock market returns as a classification problem Using TensorFlow for deep learning with Python Using deep learning for predicting stock market returns Overview slides under http://hilpisch.com/cqf_ml_elective.pdf Python Download and install Miniconda 3.6 from https://conda.io/miniconda.html If you have either Miniconda or Anaconda installed there is not need to install anything new. conda create -n elective python=3.6 (source) activate elective conda install numpy pandas=0.19 scikit-learn matplotlib conda install pandas-datareader pytables conda install ipython jupyter jupyter notebook Installing TensorFlow : https://www.tensorflow.org/install/ Finance Download the paper by Longstaff and Schwartz (2001) about the Least-Squares Monte Carlo algorithm to price American options from Paper about LSM algorithm Basic Definitions "You can think of deep learning, machine learning and artificial intelligence [AI] as a set of Russian dolls nested within each other, beginning with the smallest and working out. Deep learning is a subset of machine learning, which is a subset of AI. ... That is, all machine learning counts as AI, but not all AI counts as machine learning. For example, symbolic logic (rules engines, expert systems and knowledge graphs) as well as evolutionary algorithms and Baysian statistics could all be described as AI, and none of them are machine learning." " Neural networks are a set of algorithms, modeled loosely after the human brain, that are designed to recognize patterns. They interpret sensory data through a kind of machine perception, labeling or clustering raw input. The patterns they recognize are numerical, contained in vectors, into which all real-world data, be it images, sound, text or time series, must be translated." " Deep learning maps inputs to outputs. It finds correlations. It is known as a 'universal approximator', because it can learn to approximate the function f(x) = y between any input x and any output y, assuming they are related through correlation or causation at all. In the process of learning, a neural network finds the right f, or the correct manner of transforming x into y, whether that be f(x) = 3x + 12 or f(x) = 9x - 0.1." "All classification tasks depend upon labeled datasets; that is, humans must transfer their knowledge to the dataset in order for a neural to learn the correlation between labels and data. This is known as supervised learning. ... Any labels that humans can generate, any outcomes you care about and which correlate to data, can be used to train a neural network." Source: https://deeplearning4j.org Machine Learning Here some resources to get a first overview of basic machine learning concepts and logistic regression for classification: Visual Guide to Basics of Neural Networks Logistic Regression Basics & Algorithm Generalized Linear Models in scikit-learn Neural Networks Here some resources to get a first overview of neural networks : A Neural Network in 11 Lines of Python Introduction to Deep Neural Networks A Guide to Deep Learning Neural Networks and Deep Learning (free e-book) TensorFlow Some resources to get started with Google's TensorFlow package for deep learning: Hello, TensorFlow! TensorFlow Tutorials Data Download a HDF5 database file, containing a single pandas DataFrame object with hitorical equites data, from http://hilpisch.com/equities.h5
4. Neural Network Backbone, Layer Specifications & Loss Formulations
Palantir Technologies: Pricing, NRR, Competitive Positioning Deep Dive 1. Pricing Model Deep Dive How Palantir Prices Its Products Palantir uses a hybrid pricing model that has evolved significantly over the past two years. The structure varies by product and customer segment: Legacy Model (Gotham, pre-2023): Per-core perpetual licensing. GSA Schedule pricing (contract GS-35F-0086U) lists Palantir Gotham at $141,000 per server core for a perpetual license, plus annual support and maintenance priced per core. This was the original government pricing structure. Gotham pricing is negotiated based on: number of use cases, data volume and complexity, number of users/concurrent access, and level of customization required. No per-user licenses required; licensing is infrastructure-based. Current Model (Foundry + AIP, 2024-present): Usage-based pricing across three dimensions: Compute: Measured in compute-seconds (units of computational work), with formulas based on vCPU and GPU seconds, number of replicas, resource configuration, and time active. Storage: Measured in gigabyte-months for general-purpose data stored in non-Ontology transformation layers. Ontology: Measured in gigabyte-months (Palantir's proprietary data layer). Term licenses are also available: Palantir Foundry Term Licence requires no additional user licences, with a minimum term of 6 months. Organisation licences are priced on a per-annum basis, with included usage that must be utilized within the year (no rollover). Enterprise pricing plans offer "significant savings" over pure usage-based pricing per Palantir's documentation. UK G-Cloud 14 Pricing (public, May 2024): Palantir Platform (Foundry & AIP) is listed on the UK Digital Marketplace with: Core-based software licences priced per annum, per server core Pilot software licences priced on a term basis (scope at Palantir's discretion) Implementation and engineering services priced per person per quarter Use case maintenance plans priced per month Solution design consultancy priced per two-week sprint Specific GBP amounts are in the pricing document but were not extractable from the PDF. AWS Marketplace: Palantir Platform is available on AWS Marketplace, but pricing details are handled through private agreements rather than publicly listed rates. FedRAMP version is available as "Private Offer Only." Gotham vs. Foundry vs. AIP Pricing Differences Platform Primary Buyer Pricing Model Notes Gotham Government/defense Negotiated per-core perpetual license + annual S&M GSA: $141K/core. Highly customized. Foundry Commercial enterprise Usage-based (compute, storage, ontology) or term/org license Minimum 6-month term. No per-user fees. AIP Both Bundled with Foundry; same usage-based dimensions AIP is a capability layer on Foundry, not separately priced. Apollo Both (infrastructure) Separate pricing; available on AWS Marketplace Manages deployment and updates. AIP is not sold as a standalone product. It is a capability layer within Foundry that adds LLM orchestration, pre-built models, and data science application libraries. Customers access AIP through their Foundry license. Pricing Evolution and Shifts Pre-2023: Primarily perpetual per-core licensing (Gotham) and large annual enterprise contracts (Foundry). 2023-2024: Introduction of usage-based pricing dimensions (compute, storage, ontology) for Foundry/AIP, likely tied to the AIP launch and need to lower barriers to entry. 2024-2025: Available on AWS Marketplace and UK G-Cloud, signaling standardization. Enterprise contracts still dominate revenue, but the marketplace listing allows for more self-serve entry. No confirmed shift to outcome-based pricing as of February 2026. Land-and-Expand Pattern Initial land deals: Historically started at $1M+ for a pilot/initial deployment. AIP bootcamps are designed to compress this. Bootcamp-to-contract: ~70% of bootcamp participants convert to paid contracts within one quarter, with average initial deal size exceeding $1M (per Q1 2025 filings). Expand trajectory: A Fortune 500 industrial company expanded 5x from initial engagement. A Fortune 100 retailer converted a pilot to a $12M annual contract value within months. Top 20 customer ARPC: $54.6M (2023) -> $64.6M (2024) -> $70M (Q1 2025 TTM) -> implied higher by Q4 2025. Enterprise contracts commonly start at $10M+ annually with custom pricing based on data volume and users. Price Increases / Pricing Power No publicly disclosed standard price increases. However, the upward movement in NDR (from 107% to 139% over 2023-2025) implies significant pricing power through upsell/cross-sell. The 139% NDR means existing customers are spending 39% more year-over-year on average. The shift from per-core to usage-based pricing for Foundry/AIP may indicate Palantir is capturing more value as customers increase platform usage (natural price expansion tied to consumption). 2. Net Dollar Retention (NRR/NDR) Deep Dive Quarterly NRR Trend (2022-Q4 2025) Quarter NRR QoQ Change Notes Q4 2022 115% -- Full-year 2022 average was 115% Q1 2023 111% -4pp Decline from 2022 levels Q2 2023 ~107% -4pp Trough; coincides with AIP launch Q3 2023 107% flat Bottom Q4 2023 120% +13pp AIP-driven recovery Q1 2024 111% -9pp Methodology note: quarterly NRR can vary from TTM Q2 2024 114% +3pp Q3 2024 118% +4pp Q4 2024 120% +2pp Q1 2025 120% flat Per Q1 2025 press release Q2 2025 128% +8pp +400bps QoQ per earnings call Q3 2025 134% +6pp Q4 2025 139% +5pp Highest in company history Note on methodology: Palantir reports NDR on a trailing twelve-month basis. Some sources report slightly different figures depending on whether they reference point-in-time or TTM. The trajectory from the 107% trough in mid-2023 to 139% in Q4 2025 is unambiguous. NRR by Segment Palantir does not break out NRR by government vs. commercial in its public filings. The 139% is a blended company-wide metric. However, given: US commercial revenue grew 137% YoY in Q4 2025 US government revenue grew 66% YoY in Q4 2025 Customer count growth was faster in commercial (49% YoY to 571 US commercial) ...the commercial segment is almost certainly driving the NRR expansion more aggressively than government. What is Driving NRR Expansion Three primary drivers: AIP adoption by existing customers. AIP was launched in mid-2023 and is the primary expansion vector. Existing Foundry customers adding AIP capabilities drives incremental spend. Broader platform usage. Usage-based pricing means customers who increase compute, storage, and ontology consumption automatically spend more. Deal size expansion. TTM deals over $10M grew 108% YoY (187 deals in FY 2025 TTM). Top 20 customers ARPC rose from $54.6M to $70M+ over 2023-2025. Cohort Data Palantir does not publicly disclose cohort-level NRR data. However, management commentary suggests newer cohorts acquired through AIP bootcamps are expanding faster because they onboard with AIP from day one, versus legacy Gotham/Foundry-only customers who must cross-sell into AIP. NRR vs. Enterprise Software Peers Company NRR (Latest) Period Palantir 139% Q4 2025 Snowflake ~127% FY 2025 Databricks ~140%+ Estimated (private) CrowdStrike ~120% FY 2025 Datadog ~115% Q3 2025 ServiceNow ~120% Q4 2024 C3.ai Not disclosed -- Palantir's 139% is now among the highest in enterprise software, approaching the levels typically seen only in infrastructure/consumption companies like Snowflake and Databricks. 3. Customer Acquisition and Sales Model Sales Model Evolution Era Model Characteristics 2003-2019 Direct sales + forward-deployed engineers Long sales cycles (12-18 months). Heavy professional services. "Sales-averse" culture under Karp. Government-dominated. 2020-2022 Enterprise sales team build-out Ryan Taylor (CRO) professionalized the sales org. Foundry commercial push. Still long cycles but improving. 2023-present AIP Bootcamp flywheel 5-day intensive workshops. High-velocity deal closing. Compressed sales cycles to weeks/months. Top-of-funnel demand generation at scale. AIP Bootcamp Mechanics Duration: 5 days from zero to use case. Format: Customers bring their actual data. Palantir engineers build real workflows on that data, demonstrating AIP capabilities vs. what the customer built internally. Scale: 1,300+ bootcamps completed as of Q4 2024. Pace accelerated through 2025. Conversion rate: ~70% convert to paid contracts within one quarter. Average initial deal: Exceeds $1M. Sales cycle compression: From 1-3 months (traditional pilot) to days/weeks. Alex Karp described the bootcamp as challenging customers to compare "what they developed internally (no matter how many resources or years) to running their data on Palantir's offerings for 10 hours." Sales Efficiency Metrics Metric Q2 2025 Q3 2025 Q4 2025 CAC Payback Period 12.5 months 9.8 months 8.9 months CAC payback under 9 months is extremely efficient for enterprise software. The bootcamp model "decouples customer acquisition costs from topline growth, supporting high margins and scalable revenue expansion." Customer Count Progression Period Total Customers US Commercial YoY Growth (Total) Q4 2023 ~497 ~262 -- Q4 2024 711 ~388 43% Q1 2025 769 432 39% Q3 2025 ~882 485 ~51% Q4 2025 954 571 34% Deal Size Distribution (FY 2025 TTM) Deal Size Count YoY Growth > $1M 680 +63% > $5M 292 +90% > $10M 187 +108% Q3 2025 alone: 204 deals > $1M, 53 deals > $10M. 4. Competitive Dynamics Palantir vs. Databricks Dimension Palantir Databricks Core product AI operating system for enterprises (data integration + AI orchestration + operational workflows) Unified data lakehouse (Spark-based analytics + ML/AI) Primary buyer COO/CISO/mission owner CTO/CDO/data engineering teams Approach Top-down, embedded in operations, forward-deployed engineers Bottom-up, developer-friendly, open-source adjacent Revenue (2025) $4.5B Estimated $3B+ ARR (private, projected to exceed $3B by end 2025) Revenue growth 56% YoY (FY 2025) ~60% YoY (Q3 2024) Valuation ~$400B market cap $134B (last private round, planning IPO) Direct overlap Both offer AI/ML capabilities, data integration Databricks Data Intelligence Platform vs. Palantir AIP Differentiation End-to-end encryption, role-based controls, "trusted data fabric," government/defense pedigree Open-source ecosystem (Delta Lake, MLflow), SQL-first analytics, developer community Where they compete directly: Enterprise AI/ML deployment, data integration for large organizations. Databricks' Data Intelligence Platform is increasingly competitive with Palantir's AI capabilities. Where they are adjacent: Palantir focuses on operational decision-making and government/defense. Databricks focuses on data engineering and lakehouse analytics. Different buyers, different use cases. Palantir vs. Snowflake Dimension Palantir Snowflake Core Operational AI platform Cloud data warehouse/lakehouse Overlap Both handle large-scale data integration and analytics Snowflake "Unistore" attacking transactional + analytical workloads Differentiation Palantir is the "application layer" on top of data; Snowflake is the "storage/compute layer" Snowflake stronger in SQL analytics, concurrency, structured data Competitive dynamic Snowflake making the underlying data layer smarter, potentially reducing need for a separate "operating system" Palantir argues its Ontology layer adds intelligence that raw data infrastructure cannot Palantir vs. C3.ai Dimension Palantir C3.ai Revenue $4.5B (FY 2025) ~$400M estimated Profitability GAAP profitable, 57% adj. operating margin Sizable losses, negative FCF Model Deep integration, forward-deployed engineers, "sticky" platform 130+ pre-built turnkey AI applications, partner ecosystem (Microsoft, AWS, Google) Revenue growth 56% YoY Growing at roughly similar rate to Palantir in percentage terms (but much smaller base) GTM Direct sales + bootcamps 60%+ of deals through partner ecosystem Key difference Palantir embeds deeply and owns the customer relationship C3.ai relies heavily on hyperscaler partnerships for distribution Direct competitive encounter: C3.ai protested a SEWP award where CDC required the use of Palantir Foundry, arguing the brand-name justification was improper. Government-Specific Competitors Competitor Revenue AI Capability Competitive Dynamic Booz Allen Hamilton ~$10B+ (FY 2025) Largest AI supplier to US govt; ~$800M in AI revenue (FY 2025) Partnership with Palantir to make AI tools for logistics, autonomous systems. Complementary more than competitive. Leidos ~$16B Systems integration, AI/ML capabilities Competes on large defense programs; less software-native than Palantir SAIC ~$7.5B Implements other firms' AI tools for federal clients Low AI hiring (< 5 cleared AI positions/month Feb-Sep 2025). Not a direct AI platform competitor. CACI International ~$8B Intelligence and defense IT Competes in intelligence analytics; less commercial-scale AI Raytheon/RTX ~$80B Defense systems, some AI Hardware-focused. Palantir's software layer sits on top of RTX hardware. Palantir's structural advantage in government: From February to September 2025, Palantir and its partner ecosystem posted an average of 363-523 new security-cleared AI positions per month, far exceeding traditional contractors. This hiring velocity signals depth of AI engineering capability that legacy integrators lack. AI-Native Startup Competition Competitor Valuation Overlap Relationship with Palantir Anduril Industries $30.5B Defense hardware + AI software Strategic partner, not competitor. Part of Palantir's Warp Speed cohort. Joint work on autonomous systems. Scale AI ~$13.8B Data labeling, AI infrastructure for government Adjacent; Scale provides training data, Palantir provides operational deployment Shield AI ~$5B+ Autonomous aircraft Palantir Warp Speed cohort member; complementary Notably, Palantir has partnered rather than competed with most AI-native defense startups. The Warp Speed cohort (Anduril, L3Harris, Panasonic Energy, Shield AI) positions Palantir as the platform layer these companies build on. Win/Loss Anecdotes Win: Army DCGS protest (2016) -- Palantir protested the Army's sole-source approach, won in Federal Claims Court, and subsequently won the commercial software contract. Win: DIA Prometheus -- Palantir is currently protesting DIA's sole-source plan for Prometheus, arguing commercial software (Palantir's) is available and DIA failed adequate market research. Mirrors the DCGS playbook. Loss/Competitive pressure: Swiss government rejected Palantir at least 9 times over 7 years of attempted sales. 5. Government Business Specifics Key Contracts and Terms Contract Value Duration Status Army Enterprise Agreement Up to $10B 10 years Awarded July 2025. Consolidates 75 existing contracts into one enterprise deal. Maven Smart System (MSS) $1.3B ceiling (raised by $795M in May 2025) Through 2029 Active. AI-powered intelligence analysis software. Ceiling raised due to "growing demand." TITAN $178M initial Deliver 10 systems by 2026 Active. Tactical Intelligence Targeting Access Node -- real-time fusion of satellite, drone, and tactical vehicle data. UK MoD (Defence AI) Up to $1.5B ($750M over 5 years confirmed) Multi-year Awarded September 2025. Data analytics across classifications, interoperable with NATO. UK NHS (Federated Data Platform) $330M (GBP 330M) 7 years Awarded November 2023. UK Government Total $670M+ (GBP 670M+) cumulative Various Across at least a dozen contracts including $244M+ with NHS, $388M+ with MoD, and a contract with the nuclear weapons agency (AWE). Contract Renewal Rates Palantir does not publicly disclose contract renewal rates as a standalone metric. However: The 139% NDR implies very high renewal + expansion. The Army Enterprise Agreement (consolidating 75 contracts into one $10B vehicle) demonstrates institutional lock-in. Government contracts are "typically multi-year in nature and often expand over time as agencies add users, modules, and new operational workflows." Sole-Source vs. Competitive Bid Palantir has historically advocated for competitive procurement , filing protests when agencies attempt sole-source awards to other vendors (DIA Prometheus, Army DCGS). The company's legal strategy is to force competitions it believes it can win based on commercial software superiority. The Army Enterprise Agreement was a consolidation of existing relationships, effectively a sole-source-equivalent based on demonstrated performance. The UK MoD $240M contract was "awarded without procurement" (single-source). DOGE / Government Efficiency Exposure Direct involvement: Palantir participated in efforts with DOGE to build a "mega API" at the IRS to access data from across the agency. Alignment: The company is positioned as aligned with the Trump administration's efficiency initiatives. DOGE's mandate to modernize government IT and reduce redundancy could drive additional Palantir adoption. Risk: If DOGE results in budget cuts to agencies that are Palantir customers, spending could be affected. However, Palantir's positioning as an efficiency tool (replacing manual processes with AI) aligns with the DOGE mandate rather than being a target of cuts. Federal contract growth: Palantir's federal contracts nearly doubled in 2025, rising to $970.5M. Golden Dome SHIELD IDIQ: The Missile Defense Agency structured Golden Dome procurement through the SHIELD (Scalable Homeland Innovative Enterprise Layered Defense) multiple-award IDIQ contract. Scale: $151B total ceiling; 2,440 companies approved across three tranches (out of 2,463 applicants). Palantir's role: Positioned as "AI brain" for sensor fusion, digital twin, and AI simulation environment connecting satellites, drones, and land-based systems. Funding: $13.4B for space and missile defense in FY 2026 defense appropriations (passed Feb 3, 2026). Status: No specific contract awards announced yet as of February 2026. International Government Clients Country/Entity Contract Value Notes UK MoD, NHS, AWE (nuclear weapons) $670M+ cumulative Largest international government client. "Strategic defence AI partnership." Ukraine Battlefield deployment Not disclosed Technology "honed on the battlefield in Ukraine." Used by Ukrainian military. NATO Interoperability Included in UK MoD deal MoD systems designed to be "interoperable with NATO and other allied nations' Palantir systems." Switzerland Attempted $0 Rejected at least 9 times over 7 years. Other Various -- International revenue was ~$1.17B (26% of $4.5B FY 2025 revenue). International revenue remains the minority at 26% of total, with the US accounting for 74%. 6. TAM and Market Position Palantir's Stated TAM $1.4 trillion by 2033 (Morningstar base case, referenced in Q1 2025 earnings context). Range: Bear case $1.2T, Bull case $1.8T. This represents the total market for AI-powered operational intelligence across government and enterprise. TAM Credibility Assessment (Facts Only) The global AI software market is projected to grow from $126B (2025) to ~$400B (2030), representing a 25% CAGR. Palantir's FY 2025 revenue of $4.5B represents approximately 0.3% penetration of the stated $1.4T 2033 TAM. Against the current $126B AI software market (2025), Palantir's $4.5B represents ~3.6% share. The $1.4T TAM assumes Palantir's "ontology framework" addresses a broad market spanning defense, intelligence, healthcare, manufacturing, financial services, and other enterprise verticals. Current Market Penetration Metric Value FY 2025 Revenue $4.5B Customer Count 954 Fortune 500 penetration Not disclosed; estimated low single-digit % Government agencies served Multiple US (Army, DIA, CDC, IRS, intelligence community), UK (MoD, NHS), NATO allies % of stated TAM ~0.3% (vs. $1.4T 2033 target) % of current AI software market ~3.6% (vs. $126B 2025 market) How AI Expands/Contracts the TAM Expansion vectors: AIP makes the platform accessible to a broader set of use cases and users within existing customers. AIP bootcamps lower the barrier to entry, expanding the addressable customer base beyond traditional government/large enterprise. AI standardized templates are "reusable across industries," enabling faster deployment in new verticals. Golden Dome and other next-gen defense programs create entirely new budget categories. Contraction risk: Hyperscalers (Microsoft, AWS, Google) are making AI accessible enough that "many companies may decide they don't need Palantir's military-grade sophistication." Snowflake and Databricks are "eroding the need for a separate 'operating system' by making the underlying data layer smarter." Open-source AI frameworks could commoditize some of what Palantir offers. Key Financial Summary (Q4 2025 / FY 2025) Metric Q4 2025 FY 2025 Revenue $1.41B $4.48B Revenue Growth YoY 70% 56% US Revenue $1.08B (77% of total) ~$3.31B (74% of total) US Commercial Revenue $507M (+137% YoY) -- US Government Revenue -- (+66% YoY) ~$2.42B (54% of total) International Revenue ~$330M (23%) ~$1.17B (26% of total) Adjusted Operating Margin 57% -- Adjusted FCF $791M (56% margin) $2.27B GAAP Net Income $608M $1.63B NRR 139% -- Customer Count 954 -- TCV (quarterly) $4.3B (+138% YoY) -- Rule of 40 127% -- RPO $4.21B (+143% YoY) -- CAC Payback 8.9 months -- FY 2026 Guidance Metric Guidance Revenue $7.18-7.20B (~61% YoY growth) US Commercial Revenue >$3.14B (115%+ growth) Adjusted Operating Income $4.13B Adjusted FCF $3.9-4.1B Q1 2026 Revenue $1.53-1.54B Sources Palantir Q4 2025 Earnings Press Release Palantir Q4 2025 Investor Presentation Palantir Q3 2025 Investor Presentation Palantir Q1 2025 Financial Results CNBC: Palantir Q4 2025 Earnings BusinessWire: Palantir Q4 2025 Results Palantir Q4 2025 Earnings Highlights (MrLifeChanger) StockAnalysis: Palantir Operating Metrics DataWalk: Palantir Pricing GSA Palantir Price List UK G-Cloud 14: Palantir Platform Pricing Palantir Foundry Plans Palantir Compute Modules Usage Pricing Palantir Customer Acquisition Strategy (GTM Foundry) CNBC: Palantir $10B Army Contract DefenseScoop: Maven Contract Expansion Palantir: TITAN Contract UK MoD Palantir Deal (The Register) UK MoD Palantir Deal (Computing.co.uk) The Nerve: Palantir UK Government Contracts Washington Technology: Palantir DIA Protest The Hill: Palantir Federal Contracts FedSavvy: Palantir Federal Growth Golden Dome Guide (GovDash) Defense One: Golden Dome Competition Futurum: Palantir vs C3.ai i4C: Palantir vs Snowflake vs Databricks IO Fund: Palantir Stock Analysis ainvest: Palantir $1.2T TAM CNBC: Defense Tech Startups ExecutiveGov: Biggest Defense AI Contracts FY 2025 Palantir Blog: How AIP Bootcamps Work SaaStr: 5 Learnings from Palantir at $4B ARR Written by Hai & Claude | 2026-02-23 11:00 PST
5. Training Protocol, Convergence Behavior & Experimental Evaluation
FinRobot: An Open-Source AI Agent Platform for Financial Applications using Large Language Models FinRobot is an open-source agentic AI framework for financial decision intelligence . It combines foundation models with financial tools, quantitative models, deterministic computation, and multi-agent workflows to build decision-grade financial applications. Its core principle is simple: models reason, software computes, agents orchestrate, and systems verify. Models can be replaced; the surrounding financial infrastructure โ tools, workflows, validation, provenance, and evaluation โ is what makes the system reliable. Unlike the single-model paradigm of FinGPT , FinRobot supports end-to-end agentic workflows for equity research, quantitative trading, risk analysis, investment banking, private equity, wealth management, and other high-stakes financial decisions . Whitepaper on arXiv ยท Official Academic Page Where to start FinRobot is three projects sharing one financial domain layer. They are not three versions of the same thing competing for your attention โ they serve different purposes. Directory Version What it is Use it for finrobot_desktop/ V2 Production โ the real agent system. Native desktop research workstation on PydanticAI + FastAPI + React/Tauri Actual research work. This is the one to use if you want output you can act on finrobot_equity/ V1 Web app โ a self-hosted report generator with a FastAPI interface Standing up a browser-based service that turns a ticker into a shareable HTML/PDF report finrobot_autogen/ V0 Educational โ the original AutoGen library, the codebase behind the whitepaper Learning how financial agents are wired together, teaching, and reproducing the paper. Not intended for production use pip install finrobot installs the V0 package. Its source moved into finrobot_autogen/ , but the import name did not change โ every existing from finrobot... import ... keeps working. ๐งฌ Architecture Evolution FinRobot evolves alongside the rapid development of AI-agent frameworks. Rather than being tied to a single agent stack, each generation explores how emerging agent architectures can improve financial analysis, research, and decision-making. Version Agent Framework Project Maturity Focus V0 AutoGen finrobot_autogen/ Educational / reference The original FinRobot multi-agent architecture for financial applications V1 OpenAI Agents SDK finrobot_equity/ Self-hosted web app Equity-research agents, financial analysis, valuation, and automated report generation V2 PydanticAI finrobot_desktop/ Production โ also hosted at finrobot.ai/v2 Structured, type-safe agent workflows for professional equity research V3 DeepSeek-Harness FinRobot V3 In development More autonomy inside the same verification constraints All three are open source. V2 is the production system โ the deterministic compute engine, the provenance guarantees, and the agent orchestration described below all live there. V0 is kept because it is small enough to read and learn from, not because it is the recommended way to run research today. Our philosophy: FinRobot is not defined by any single agent framework. We continuously adopt, evaluate, and evolve with state-of-the-art agent architectures while keeping the financial domain layer โ tools, workflows, deterministic computation, and decision-making capabilities โ at the core. ๐งฎ Deterministic compute, LLM narration A design principle that runs through every generation: deterministic financial computation and LLM-based narration stay strictly separate. All financial numbers are produced by pure-Python compute operators, not by the language model. The LLM handles reasoning, synthesis, explanation, and report writing, while valuation outputs โ DCF, DDM, LBO, WACC, comparable-company analysis, Monte Carlo simulation โ are calculated through deterministic code paths with full provenance. Numbers are code-calculated. Narratives are LLM-assisted. Every output is provenance-tracked. ๐ FinRobot Desktop v0.1.0 A native desktop equity research cockpit powered by a production-grade multi-agent architecture. It takes analysts from market data and company filings to valuation, debate, synthesis, and investment-committee-style reports in one traceable workflow. Stock workspace โ live market data pulled independently of any AI report, beside the research verdict and the DCF / DDM / LBO / comps instruments. Research report โ 13 chapters, a 12-month target with its range, model confidence, and the data caveats the audit operators raised. The design principle, visible in the product: the DCF table is labelled code-computed โ a ten-year forecast from pure-Python operators โ while the paragraph beside it is the LLM reading those numbers back across all three valuation methods. ๐ Latest release: FinRobot Desktop v0.1.0 For macOS Apple Silicon (M1/M2/M3 or later), download FinRobot_0.1.0_aarch64.dmg and drag FinRobot into Applications . Intel Mac builds are not available in this release. The app is not yet Apple-notarized, so on first launch macOS may report that it is "damaged." Run this once in Terminal, then open it normally: xattr -cr /Applications/FinRobot.app Multi-agent architecture A Lead Agent orchestrates specialized research agents through a pipeline-driven execution engine: five role-based sub-agents for data, analysis, modeling, synthesis, and reporting, plus three debate agents for the bull case, bear case, and judgment. User Research Request โ Lead Agent / Orchestrator โ Data Agent โ Analysis Agent โ Modeling Agent โ Synthesis Agent โ Report Agent โ Bull Agent โ Bear Agent โ Judge Agent โ Traceable Investment Research Output Codebase snapshot Layer What it includes Full-stack system ~184k lines across a Python backend, React/Tauri desktop frontend, Rust shell, and tests Agent runtime 9 agents: lead orchestrator, 5 role-based pipeline agents, 3 debate agents Research pipelines 7 pipelines โ company research, DCF, comps, LBO, DDM, earnings, IC memo Deterministic compute 32 pure-Python operators (26 valuation/analysis + 6 audit) and 7 coordinators Data infrastructure 7 providers with failover โ FMP, Finnhub, yfinance, SEC EDGAR, Adanos, NewsAggregator, FX Skills 56 analyst playbooks across equity research, investment banking, private equity, and wealth management Product stack PydanticAI, FastAPI, SQLite, React 19, Vite 6, Zustand, Tauri/Rust, Recharts Full details in finrobot_desktop/README.md . Getting started V2 โ the desktop app, a local web UI, or the CLI Three ways to run the same engine. Desktop app โ download the release (macOS Apple Silicon). Local web UI โ the same interface in a browser, no .dmg and no Rust toolchain, so it also works on Intel Macs, Linux, and Windows: cd finrobot_desktop uv sync # backend dependencies (cd desktop && npm install) # frontend dependencies โ one time ./dev.sh # โ open http://localhost:5173 dev.sh runs the FastAPI backend on :8321 and a Vite server on :5173 that proxies the API to it; Ctrl+C stops both. Note that it first frees those two ports, so quit FinRobot.app if it is open, and that the local API is unauthenticated in browser mode โ details in finrobot_desktop/README.md . CLI โ after uv sync in finrobot_desktop/ : finrobot research AAPL # full 13-chapter research artifact finrobot dcf MSFT # DCF valuation (auto-switches to DDM where appropriate) finrobot comps NVDA --peers AMD,INTC finrobot ic-memo TSLA finrobot ask AAPL " How exposed is the gross margin to tariffs? " Building the Tauri desktop shell is covered in finrobot_desktop/README.md . V1 โ the equity research web app cp finrobot_equity/core/config/config.ini.example finrobot_equity/core/config/config.ini # edit config.ini: fmp_api_key, openai_api_key, (optional) adanos_api_key chmod +x finrobot_equity/deploy.sh ./finrobot_equity/deploy.sh start # โ http://127.0.0.1:8001 Command Description ./finrobot_equity/deploy.sh start Start the web app (auto-installs dependencies) ./finrobot_equity/deploy.sh stop Stop the application ./finrobot_equity/deploy.sh restart Restart the application ./finrobot_equity/deploy.sh status Check running status ./finrobot_equity/deploy.sh install Install/update dependencies only If deploy.sh doesn't work in your environment: python3 -m venv venv && source venv/bin/activate pip install -r finrobot_equity/requirements.txt python finrobot_equity/run_web_app.py A two-step CLI pipeline is available as well โ see finrobot_equity/README.md . Example reports: NVDA ยท MSFT ยท TSLA ยท META ยท COP V0 โ the AutoGen framework 1. Create an environment (Python 3.10 or 3.11): conda create --name finrobot python=3.10 conda activate finrobot 2. Install โ from PyPI, or from source at the repository root ( setup.py maps the finrobot package to finrobot_autogen/finrobot , so installing from inside that directory won't work): git clone https://github.com/AI4Finance-Foundation/FinRobot.git cd FinRobot pip install -e . # or: pip install -U finrobot 3. Configure keys โ both files go in finrobot_autogen/ , which is where the notebooks look for them. Copy rather than rename: the *_sample files are tracked, and your filled-in copies are gitignored. cd finrobot_autogen cp OAI_CONFIG_LIST_sample OAI_CONFIG_LIST # OpenAI / Azure OpenAI endpoints cp config_api_keys_sample config_api_keys # Finnhub, FMP, SEC, Reddit, โฆ 4. Run a tutorial from finrobot_autogen/tutorials_beginner/ or tutorials_advanced/ : agent_annual_report.ipynb # 10-K โ formatted PDF annual report agent_fingpt_forecaster.ipynb # market forecast from news + financials agent_trade_strategist.ipynb # strategy writing and backtesting lmm_agent_mplfinance.ipynb # multimodal agent reading a candlestick chart lmm_agent_opt_smacross.ipynb # multimodal SMA-crossover tuning The agent library, workflow types, and full tutorial index are in finrobot_autogen/README.md . Repository layout FinRobot/ โโโ finrobot_autogen/ # V0 โ AutoGen generation (PyPI: pip install finrobot) โ โโโ finrobot/ # package root โ imported as `finrobot` โ โ โโโ agents/ # agent_library.py, workflow.py, prompts.py โ โ โโโ data_source/ # finnhub / finnlp / fmp / sec / yfinance / reddit โ โ โโโ functional/ # analyzer, charting, coding, quantitative, rag, text โ โ โโโ toolkits.py # registers Python functions as agent tools โ โ โโโ utils.py โ โโโ tutorials_beginner/ # hands-on tutorials โ โโโ tutorials_advanced/ # advanced tutorials for FinRobot developers โ โโโ experiments/ # investment group, multi-factor, portfolio optimization โ โโโ configs/ report/ # agent configs and sample generated reports โ โโโ FinNLP/ # git submodule โ โโโ OAI_CONFIG_LIST_sample โ โโโ config_api_keys_sample โ โโโ requirements.txt โ โโโ finrobot_desktop/ # V2 โ PydanticAI desktop generation (current) โ โโโ finrobot/ # Python backend (FastAPI + compute engine) โ โ โโโ engine/ # agents/, pipelines/, compute/, data/ โ โ โโโ artifact/ # report store + output contract gate โ โ โโโ audit/ coverage/ obs/ routes/ โ โ โโโ cli.py server.py sdk.py โ โโโ desktop/ # Tauri shell + React frontend (src/, src-tauri/) โ โโโ skills/ # 56 analyst playbooks โ โโโ tests/ scripts/ tutorials/ โ โโโ pyproject.toml uv.lock dev.sh โ โโโ finrobot_equity/ # V1 โ OpenAI Agents SDK generation โ โโโ core/ # analysis engine + 8 section-writing agents โ โโโ web_app/ # FastAPI web application โ โโโ run_web_app.py # launcher โ โโโ deploy.sh # local deployment โ โโโ deploy.gcloud.sh # Cloud Run deployment โ โโโ requirements.txt โ โโโ Dockerfile .dockerignore # V1 container build โ must stay at the repo โ # root, since the image imports the app as โ # finrobot_equity.web_app.main โโโ .github/workflows/ # desktop CI (backend 3.11/3.12, frontend Node 22/24) โโโ setup.py # packages V0 (as `finrobot`) + V1 for PyPI โโโ LICENSE NOTICE TRADEMARK_POLICY.md โโโ README.md ๐ฌ FinRobot Pro โ your personal AI-powered equity research assistant ๐ https://finrobot.ai/ โถ๏ธ Click the image above to watch the demo video, or see the short preview below. FinRobot-Intro.mp4 FinRobot Pro is an AI-powered equity research platform that automates professional stock analysis using large language models and AI agents: Automated report generation โ professional equity research reports on demand Financial analysis โ income statements, balance sheets, and cash flows Valuation analysis โ P/E, EV/EBITDA multiples, and peer comparison Risk assessment โ comprehensive investment risk evaluation Research foundation The FinRobot ecosystem The overall framework is organized into four layers, each addressing a specific aspect of financial AI processing: Financial AI Agents Layer โ includes Financial Chain-of-Thought (CoT) prompting to strengthen complex analysis and decision-making. Market Forecasting Agents, Document Analysis Agents, and Trading Strategies Agents use CoT to break financial problems into logical steps, aligning their algorithms and domain expertise with evolving market dynamics. Financial LLMs Algorithms Layer โ configures and applies models tuned to specific domains and global market analysis. LLMOps and DataOps Layers โ a multi-source integration strategy that selects the most suitable LLM for each financial task across a range of state-of-the-art models. Multi-source LLM Foundation Models Layer โ supports plug-and-play use of general and specialized LLMs. Agent workflow Perception โ captures and interprets multimodal financial data from market feeds, news, and economic indicators, structuring it for analysis. Brain โ the core processing unit: consumes perception output with LLMs and applies Financial CoT to generate structured instructions. Action โ executes those instructions with tools, turning analysis into outcomes: trades, portfolio adjustments, reports, or alerts. Smart Scheduler The Smart Scheduler ensures model diversity and selects the most appropriate LLM for each task. Director Agent โ orchestrates task assignment, allocating work based on performance metrics and task suitability. Agent Registration โ manages registration and tracks agent availability for efficient allocation. Agent Adaptor โ tailors agent functionality to specific tasks. Task Manager โ stores and manages general and fine-tuned LLM-based agents for different financial tasks, updated periodically. AI Agent papers [Stanford University + Microsoft Research] Agent AI: Surveying the Horizons of Multimodal Interaction [Stanford University] Generative Agents: Interactive Simulacra of Human Behavior [Fudan NLP Group] The Rise and Potential of Large Language Model Based Agents: A Survey [Fudan NLP Group] LLM-Agent-Paper-List [Tsinghua University] Large Language Models Empowered Agent-based Modeling and Simulation: A Survey and Perspectives [Renmin University] A Survey on Large Language Model-based Autonomous Agents [Nanyang Technological University] FinAgent: A Multimodal Foundation Agent for Financial Trading: Tool-Augmented, Diversified, and Generalist AI Agent open-source frameworks & tools AutoGPT (183k stars) : autonomous AI agent platform. Dify (134k stars) : LLM app development platform with workflow orchestration and RAG. LangChain (130k stars) : framework for building context-aware LLM applications. MetaGPT (65.6k stars) : multi-agent framework with role-based collaboration. AutoGen (56k stars) : framework for multi-agent LLM applications with tools and human interaction. CrewAI (46.6k stars) : framework for orchestrating collaborative AI agents. ChatDev (31.7k stars) : multi-agent framework for software development tasks. FastGPT (27.4k stars) : knowledge-based LLM platform with workflow support. Langfuse (23.4k stars) : open-source LLM observability and evaluation platform. BabyAGI (22.2k stars) : task-driven experimental autonomous agent framework. SuperAGI (17.3k stars) : developer-focused autonomous agent framework. CAMEL (16.4k stars) : framework for cooperative and communicative AI agents. Bisheng (11.2k stars) : enterprise open-source LLM application platform. Citing FinRobot @article { yang2024finrobot , title = { FinRobot: An Open-Source AI Agent Platform for Financial Applications using Large Language Models } , author = { Yang, Hongyang and Zhang, Boyu and Wang, Neng and Guo, Cheng and Zhang, Xiaoli and Lin, Likun and Wang, Junlin and Zhou, Tianyu and Guan, Mao and Zhang, Runjia and Wang, Christina Dan } , journal = { arXiv preprint arXiv:2405.14767 } , year = { 2024 } , doi = { 10.48550/arXiv.2405.14767 } , url = { https://arxiv.org/abs/2405.14767 } } @inproceedings { zhou2024finrobot , title = { FinRobot: {AI} Agent for Equity Research and Valuation with Large Language Models } , author = { Tianyu Zhou and Pinqiao Wang and Yilin Wu and Hongyang Yang } , booktitle = { ICAIF 2024: The 1st Workshop on Large Language Models and Generative AI for Finance } , year = { 2024 } } @inproceedings { han2024enhancing , title = { Enhancing Investment Analysis: Optimizing AI-Agent Collaboration in Financial Research } , author = { Han, Xuewen and Wang, Neng and Che, Shangkun and Yang, Hongyang and Zhang, Kunpeng and Xu, Sean Xin } , booktitle = { ICAIF 2024: Proceedings of the 5th ACM International Conference on AI in Finance } , pages = { 538--546 } , year = { 2024 } } License Apache 2.0 โ see LICENSE . Trademark usage is covered by TRADEMARK_POLICY.md . Disclaimer : The code and documents provided here are released under the Apache-2.0 license. They should not be construed as financial advice or recommendations for live trading. Exercise caution and consult qualified financial professionals before any trading or investment decisions.
Model Figures & Evaluation Curves (4 Images)
Deep Learning Execution Pipeline
(Preprocessing)
(Feature Extraction)
(Self-Attention)
(Loss Optimization)