Investment OS

A local-first AI investment analysis platform that combines Python, broker APIs, Claude, and MCP to provide personalized portfolio insights while keeping credentials and sensitive data local.

InvestmentOS is a local-first AI-powered investment analysis platform built to explore how AI can be used to understand and analyze personal investment data through natural language.

The project started with a simple idea: what if an AI-powered financial consultant could understand my portfolio, remember my investment goals and investment philosophy, and help me analyze my investments through conversation?

Privacy was a key design requirement. Broker credentials are kept on the local machine and are never directly exposed to Claude through MCP. InvestmentOS retrieves and processes portfolio data locally, generates structured reports and historical snapshots, and Claude accesses only the approved generated data through Filesystem MCP.

An investment context containing goals, risk profile, and investment preferences allows the AI to provide more personalized analysis rather than generic responses.

The architecture combines Python, broker APIs, Claude AI, Claude Dispatch, Filesystem MCP, and Bash automation to create an AI-assisted investment analysis workflow.

The project is an ongoing exploration of AI orchestration, MCP, local-first architecture, privacy, and conversational interfaces for real-world applications.