Investing Companion
Self-hosted equity analysis with AI-powered insights.
Dual-stack dashboard combining stock data from multiple sources with Claude AI analysis, TradingView charts, and Discord alerts. Designed for personal investment research without subscription fees.
Demo credentials
Email: demo@example.com Password: demo1234!
Read-only mode — data resets weekly. Access restricted to US/Canada.
Tech Stack
Why I Built This
Premium stock analysis tools cost $30-50/month and still don't let you customize the analysis workflow. I wanted a personal dashboard where I could combine data from multiple sources, run my own AI analysis, and get alerts on my terms.
Features
- • Equity dashboard with real-time quotes and interactive TradingView charts
- • Fundamental analysis with 20+ financial ratios and peer comparison
- • Claude-powered AI analysis considering price history and portfolio context
- • Watchlists with custom columns and cross-equity comparison
- • Configurable price alerts (crosses, percent change, volume spikes) via Discord
- • Trade tracker with P&L calculations and position history
- • Market overview with index tracking and sector heatmaps
- • Earnings calendar and macro economic event tracking
Architecture
The backend is a FastAPI service handling data collection, Celery tasks, and AI analysis. The frontend is a Next.js app with TanStack Query for data fetching. TimescaleDB handles time-series price data efficiently, and Redis serves as the Celery broker and cache layer.
Key Decisions
Dual-stack (Python + TypeScript)
Python's ecosystem for financial data (pandas, yfinance) is unmatched. Next.js provides the best React DX. The two communicate via REST API.
TimescaleDB over plain PostgreSQL
Hypertables with automatic chunking and compression make time-series queries fast without manual partitioning.
Claude API for analysis
Instead of building custom scoring models, Claude can synthesize fundamentals, price trends, and news into coherent analysis. The prompts are tuned for concise, actionable output.