Release: V0.1.0
Repository binding:
- Eduvane AI - Bound to Product Definition v1.2
- Architectural Policy Compliant
- Detova Labs Implementation
- Frontend (React + TypeScript):
apps/web - Backend Gateway (Node + TypeScript):
apps/gateway - AI Engine (Python):
apps/ai-engine - Shared contracts:
packages/contracts
The frontend never performs intent orchestration. All intelligence routing is handled by the Python AI engine through the Node gateway.
Existing root-level prototype files are preserved for reference. New development should target apps/web, apps/gateway, and apps/ai-engine.
If you still run the root prototype, use services/.env.example as its environment template.
Prerequisites:
- Node.js
20+(recommended: use.nvmrc) - npm
10+ - Python
3.12+withvenv
- Copy environment templates:
apps/web/.env.example->apps/web/.envapps/gateway/.env.example->apps/gateway/.envapps/ai-engine/.env.example->apps/ai-engine/.env
- Install dependencies:
npm installnpm --prefix apps/web installnpm --prefix apps/gateway installnpm --prefix packages/contracts install
- Install Python dependencies:
python3 -m venv apps/ai-engine/.venvsource apps/ai-engine/.venv/bin/activatepip install -r apps/ai-engine/requirements.txt
- Run services in separate terminals:
- Web:
npm --prefix apps/web run dev - Gateway:
npm --prefix apps/gateway run dev - AI Engine:
apps/ai-engine/.venv/bin/uvicorn app.main:app --app-dir apps/ai-engine --reload --port 8090
- Web:
apps/web:- Chat-first workspace UI
- Bottom-anchored composer with integrated upload control
- Sidebar for conversation history
apps/gateway:- Supabase session validation
- Input validation and request shaping
- Security middleware and AI engine forwarding
apps/ai-engine:- Intent detection and orchestration
- Role-aware response synthesis
- Handwriting quality feedback hooks
- Session memory per chat session