This project offers an open-source, AI-powered financial assistant, democratizing access to institutional-grade data and analysis tools. It enables complex queries, code execution, and visualizations through natural language, bridging the gap between expensive financial terminals and accessible AI interfaces, powered by the Valyu API.

How It Works

Finance integrates the Valyu API for comprehensive financial data (markets, filings, research). It uses LLMs (OpenAI or local via Ollama/LM Studio) for natural language understanding and Daytona sandboxes for secure Python code execution, supporting ML models and backtesting. Interactive visualizations are generated using Recharts. The architecture supports cloud-based production and fully local development modes, prioritizing privacy and cost-efficiency with local LLM options.

Quick Start & Requirements

Two modes are available: Development (recommended for forking) and Production.

• Development Mode:

  • Install: git clone, cd finance, npm install.
  • Prerequisites: Node.js 18+, npm/yarn, Valyu API key, Daytona API key. Optional: Ollama or LM Studio for local LLMs.
  • Setup: Configure .env.local with API keys.
  • Run: npm run dev (access http://localhost:3000).
  • Benefits: No Supabase/auth, local SQLite, unlimited queries, offline capability.
  • Docs: Live demo at finance.valyu.ai, detailed guides in repo.

• Production Mode: Requires Node.js 18+, npm/yarn, Valyu credentials, OpenAI API key, Daytona API key, and Supabase setup. Deployment typically via Vercel.

Highlighted Details

• Access to live global market data, SEC filings, patent databases, academic research, and web search via Valyu API.
• Secure Python code execution in Daytona sandboxes for ML, backtesting, and custom financial models.
• Generation of interactive, publication-ready charts and dashboards.
• Support for local LLMs (Ollama, LM Studio) for private, unlimited queries, with OpenAI as fallback.
• Unified data access, replacing fragmented, expensive platforms.

Maintenance & Community

Developed by the Valyu team, with links to Twitter, LinkedIn, and GitHub. Contributions are welcomed via Pull Requests.

Licensing & Compatibility

Licensed under the MIT License, permitting commercial use and integration into closed-source projects.

Limitations & Caveats

Production deployment requires significant setup across multiple services (Valyu, Supabase, OpenAI, Daytona). Using local LLMs via LM Studio necessitates configuring a large context window (8192+ tokens) per model to prevent errors during complex tool interactions. Development mode uses local SQLite, separate from production Supabase data.