DevFlow AI is a unified, AI-powered developer workspace designed to completely eliminate context-switching.
It brings project planning, intelligent codebase analysis, real-time RAG (Retrieval-Augmented Generation), automated debugging, and persistent developer notes into a single cohesive platform.
Most importantly, DevFlow AI is powered by locally-running Large Language Models (LLMs) via Ollama. Your proprietary code never leaves your machine. No API costs. Total privacy.
DevFlow AI is built as a robust Monorepo using the latest enterprise-grade technologies.
| Layer | Technology | Description |
|---|---|---|
| Web App | Next.js 15, React 19 (Hosted on Vercel) | App Router, Server Components, shadcn/ui components |
| Mobile App | React Native CLI | Native iOS & Android application sharing the same core backend |
| Backend API | NestJS (Hosted on Render) | Modular architecture, WebSockets for real-time progress |
| Database | PostgreSQL + pgvector (Hosted on Neon.tech) | Relational data + Vector embeddings for semantic AI search |
| Cache & Auth | Redis (Hosted on Upstash) | JWT Token blacklisting and session management |
| AI Runtime | Ollama (Hosted on Oracle Cloud VM) | Remote execution of Llama 3 and Nomic Embeddings |
| ORM | Prisma 7 | Type-safe database interactions and automated migrations |
| State | Zustand + React Query | Global UI state management and server-data caching |
- 🔒 Stateless & Secure Auth: JWT-based authentication with seamless background token refreshing and Redis-backed logout blacklisting.
- 📂 Workspaces & Projects: Organize your codebase hierarchy to keep mental context clean.
- 📝 Markdown Developer Notes: Live-preview markdown editor with auto-save and syntax highlighting.
- 🧠 Context-Aware AI Chat: Ask questions about your code. The backend uses
pgvectorto perform semantic similarity searches across your indexed repositories to provide pinpoint accurate answers. - ⚡ Real-Time Streaming: AI responses are streamed token-by-token to both Web and Mobile via Server-Sent Events (SSE).
- 📱 Cross-Platform Sync: The React Native mobile app shares the exact same backend and database, meaning your notes and AI chats are available on the go.
devflow-ai/
├── apps/
│ ├── web/ # Next.js Web Frontend (Port 3000)
│ ├── backend/ # NestJS API Server (Port 3001)
│ └── mobile/ # React Native iOS & Android App
├── documents/ # Detailed Engineering Architecture Guides
├── docker-compose.yml# Local Infrastructure (Postgres, Redis, Ollama)
└── package.json # Root workspace configuration
- Node.js 20+
- Docker Desktop (with WSL 2 enabled on Windows)
- Git
We use Docker to instantly spin up PostgreSQL (with pgvector), Redis, and Ollama.
docker compose up -dYou need an LLM for chat and an embedding model for vectorizing your code.
docker exec -it devflow_ollama ollama pull llama3
docker exec -it devflow_ollama ollama pull nomic-embed-textcd apps/backend
cp .env.example .env # Ensure variables map to the docker ports
npm install
npx prisma migrate dev
npx prisma db seed # Optional: populate test data
npm run start:devIn a new terminal:
cd apps/web
npm install
npm run devOpen http://localhost:3000 to view the web application.
In a new terminal:
cd apps/mobile/DevFlowMobile
npm install
npm start -- --reset-cache
### 6. Run E2E Tests
We use Playwright for robust End-to-End testing of our core web flows. Make sure both the backend and web frontend are running locally, then execute:
```bash
cd apps/web
npm run test:e2e
This will run the test suite in headless Chromium mode and output a detailed HTML report.
For an exhaustive, deep-dive look into how this system was built, please refer to the markdown guides located in the /documents folder:
DevFlowAI-Documentation_1.md: Weeks 1-3 (Foundation & Auth)DevFlowAI-Documentation_2.md: Weeks 4-10 (RAG, WebSockets, GitHub Indexing)DevFlowAI-Documentation_3.md: Weeks 11-12 (Mobile App & Polish)DevFlowAI-Documentation_4.md: Weeks 13-14 (Production Deployment, Neon, Render, Oracle Cloud, E2E & Load Testing)DevFlowAI-Documentation_5_Architecture.md: (Complete System Architecture & Cloud Infrastructure Diagrams)
This project is licensed under the MIT License.