AI Stock Assist

AI-powered stock analysis in seconds. The disciplined frameworks of Buffett, Graham, Munger, Lynch, Bogle, and Housel — applied to real-time market data, in plain English.

Live product: aistockassist.com

---

🏗️ Architecture (production)

Browser
  │
  ▼
aistockassist.com  ◄── Vercel (Static React SPA — this repo)
  │
  │  fetch
  ▼
api.aistockassist.com  ◄── DNS CNAME
  │
  ▼
ai-stock-render-api.onrender.com  ◄── Render (FastAPI — ../ai-stock-render repo)
  │
  ├─► yfinance         (market data)
  ├─► Google Gemini    (AI recommendations, with Groq Llama fallback)
  └─► Supabase         (auth + analysis history + user profiles)
                              │
                              ▼
                         Postgres (Supabase managed)

Payments: Stripe → Vercel serverless webhook → Supabase

Repo layout

Repo	Hosts	Purpose
ai-stock-assist-web (this repo)	Vercel	React 19 + TypeScript + Vite 6 + Tailwind v4 SPA
ai-stock-render (sibling)	Render service ai-stock-render-api	FastAPI backend serving /api/analyze, /api/recommendation, etc.

⚠️ Service name vs DNS: Note that render.yaml in the sibling repo defines a service named ai-stock-assist-api, but the live Render service serving api.aistockassist.com is named ai-stock-render-api. They are different services from the same codebase. See docs/PROJECT-CLEANUP-PLAN-JUNE-2026.md for the consolidation plan.

---

🕰️ Architecture history

Generation	Stack	Status
Gen 1 (early 2025)	Streamlit single-file app	❌ Superseded
Gen 2 (mid 2025–Q1 2026)	Streamlit + Supabase auth + Stripe + AI services. Polished MVP.	❌ Superseded by Gen 3, slated for deletion in June 2026 cleanup
Gen 3 (Q1 2026 →)	React SPA on Vercel + FastAPI on Render + Supabase. Current production.	✅ Live

The legacy Streamlit service still runs at ai-stock-assist.onrender.com only because nobody's flipped the off switch yet. It has no role in the live product. See cleanup plan above.

---

🚀 Development

Prerequisites

• Node 20+ (frontend)
• Python 3.11+ (backend — sibling repo)
• Supabase project, Stripe account, Google Gemini + Groq API keys

Frontend (this repo)

# Install
npm install

# Run dev server (Vite on :5173)
npm run dev

# Type check + build
npm run build

# Preview production build
npm run preview

Environment variables for the frontend live in .env.local. Canonical record is .env.example. See DEPLOYMENT.md for the full list.

Backend (sibling repo ai-stock-render)

cd ../ai-stock-render/api
pip install -r requirements.txt
uvicorn server:app --reload --port 8000

Set the frontend's VITE_API_URL=http://localhost:8000 in .env.local to point at the local backend.

---

🚢 Deployment

Both apps auto-deploy on push to main:

Push to	Deploys to	Time
ai-stock-assist-web main	Vercel production	~90s
ai-stock-render main	ai-stock-render-api on Render	~2–5 min

⚠️ Testing usually happens in production (no .env.local configured for most contributors). Always add Vercel / Render env vars before pushing code that depends on them.

For full deployment runbook: DEPLOYMENT.md. For deployment sequencing & lessons: CLAUDE-LESSONS-LEARNED.md.

---

📚 Documentation

File	Purpose
DEPLOYMENT.md	Step-by-step deployment runbook, env var checklist
CLAUDE-LESSONS-LEARNED.md	Bug catalog, gotchas, "don't do this" list — read before touching anything load-bearing
docs/RELEASE-NOTES-MAY-2026.md	Latest release notes
docs/PROJECT-CLEANUP-PLAN-JUNE-2026.md	Streamlit + orphan-service decommission plan
docs/PUBLISHED-CONTENT-TRACKER.md	Marketing campaign log (Substack, LinkedIn, X, FB URLs)
docs/PUBLISHING-PLAN-MAY-2026-LIBRARY-AND-EP5.md	Active publishing roadmap
docs/CLAUDE-AGENT-BRIEFING-TEMPLATE.md	Template for handing off blog-episode prep to a fresh Claude session

---

🤝 Built with

• React 19 + TypeScript + Vite 6
• Tailwind v4 + custom CSS variables (src/index.css)
• Supabase (auth, postgres, real-time)
• Stripe (credit-pack payments)
• FastAPI + yfinance + Google Gemini + Groq Llama (backend)
• Vercel (frontend hosting) + Render (backend hosting)

---

📜 License

© 2026 AI Stock Assist · Lindsay Hiebert · Part of the AI for Good initiative under PI GenAI.

Disclaimer: AI Stock Assist provides AI-generated analysis for educational purposes only. This is not financial advice. Always do your own research before making investment decisions. Past performance does not guarantee future results.