AI-powered business intelligence platform for Vietnamese restaurant operators. We help chain owners and skilled operators make data-driven business decisions through AI recommendations.
This repository contains all strategic planning, architecture, and design documents for the project.
We're not a menu builder—we're an AI business manager. Menu creation and online ordering are data collection tools that feed our AI models to deliver actionable business recommendations.
Segment A: Chain owners (3-10 locations) who lack TIME to manage all locations
Segment B: Skilled operators (chefs/baristas) who lack BUSINESS KNOWLEDGE
Unified Value Proposition: "AI quản lý kinh doanh cho chủ quán không có thời gian hoặc kiến thức về kinh tế"
- AI Analytics Dashboard (80%) - Core product delivering business insights
- Data Collection Tools (15%) - Menu, ordering, POS integration
- Customer-Facing Tools (5%) - Public storefronts, QR menus
Market: Vietnam only (Phases 1-3). Deep integrations with MoMo, ZaloPay, GrabFood, Zalo.
📖 For complete strategy, see planning/product-strategy-refined.md
- /planning: Contains the feature roadmap and monetization strategy.
- /research: Contains market analysis and go-to-market strategies for Vietnam and Japan.
- /design: Contains wireframes, UX flows, and UI mockups.
- /architecture: Contains the system architecture diagram and technology stack decisions.
- /knowledge-base: Living playbooks for market-specific execution (currently focused on Vietnam).
Developer artifacts (OpenAPI for codegen, seeds, CI) live in the implementation repository — see documentation/implementation-repo-README-template.md for expected layout.
- Mobile (Owner): Flutter (iOS + Android) - Real-time dashboard, push notifications
- Web Portal (Staff): Next.js + Tailwind CSS - Admin operations, data entry
- API Gateway: NestJS (TypeScript) - GraphQL, WebSocket, REST
- AI Service: Python + FastAPI - ML models, recommendations engine
- Communication: gRPC (backend ↔ AI), GraphQL + WebSocket (client ↔ backend)
- Database: PostgreSQL 14+ (multi-tenant with Row-Level Security)
- Cache: Redis 6+ (metrics, sessions, pub/sub)
- Storage: S3-compatible (images, exports)
- Deployment: Docker + Docker Compose (dev), Kubernetes (prod)
📖 See architecture/system-diagram.md for complete architecture details
Quick Start (30 minutes):
# Prerequisites: Node.js 20+, Python 3.11+, Docker
# 1. Start database services
docker-compose up -d
# 2. Setup backend (NestJS)
cd backend
npm install
npm run start:dev
# 3. Setup AI service (Python)
cd ../ai-service
python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt
python -m app.grpc_server
# 4. Setup mobile app (Flutter)
cd ../mobile
flutter pub get
flutter run📖 Full setup guide: QUICK-START.md
- Backend Setup: architecture/backend-setup-guide.md
- GraphQL Schema: architecture/graphql-schema.md
- Flutter App Spec: architecture/flutter-mobile-app-spec.md
- Sprint 1 Plan: planning/sprint-1-implementation.md (4 weeks to MVP)
Current Status: Implementation Phase - Sprint 1 (Oct 22 - Nov 19, 2025)
Sprint 1 Deliverables:
- ✅ Phone OTP authentication
- ✅ Real-time dashboard (revenue, orders, AI recommendations)
- ✅ Push notifications (iOS + Android)
- ✅ Multi-location support
- ✅ AI recommendation engine (rule-based)