Backend (Spring Boot, Kafka, Redis, Gemini) and frontend (React, Tailwind) for an AI-powered email draft assistant. Connects to Gmail, categorizes incoming mail, generates draft replies with AI, and lets you review or skip before sending.

• backend/ – Spring Boot app (Gmail → Kafka → categorize → draft → Redis). Run from here.
• frontend/ – Vite + React + Tailwind dashboard. Run from here.
• docker/ – Kafka, Redis, Kafka UI, RedisInsight (Compose).
• test/ – Demo email seed script for local/demo use.

• Java 21, Maven
• Node.js 18+ (for frontend)
• Docker (for Kafka + Redis)
• Google Cloud project with Gmail API and OAuth credentials
• Gemini API key (for categorization and draft generation)

1. Start infrastructure:

   docker compose -f docker/docker-compose.yml up -d

2. Configure backend: copy backend/src/main/resources/application-local.yml.example to application-local.yml and set OAuth client ID/secret, Gemini API key, etc.

3. Run backend:

   cd backend && mvn spring-boot:run

   App listens on port 8080.

4. Run frontend:

   cd frontend && npm install && npm run dev

   Open http://localhost:5173. The dev server proxies /api to the backend.

5. Log in via the app (OAuth with Gmail), then use Drafts or seed demo data (see below).

To trigger the pipeline with fake emails for a demo:

./test/seed-demo-emails.sh

Uses test/demo-emails.json and POST /api/gmail/seed. Backend and Kafka must be running.

• Code: see LICENSE.
• Docs: LICENSE-DOCS.md (CC-BY 4.0).