🎓 Meri AI

AI-Powered Campus Navigation & Intelligence System Adama Science and Technology University

⭐ Star · 🐞 Report Bug · ✨ Request Feature

🎯 About

Meri AI is an intelligent campus navigation and knowledge management system built specifically for Adama Science and Technology University (ASTU).

It combines AI-powered conversational assistance, interactive campus mapping, and retrieval-augmented generation (RAG) to help students, staff, and visitors navigate the campus and access accurate information effortlessly.

💡 Why Meri AI?

🤖 AI-Powered Assistant — Natural language queries using Google Gemini

🗺️ Smart Navigation — Real-time routing powered by OpenStreetMap

📚 Knowledge Base — Semantic campus search with RAG

📱 Progressive Web App — Installable, offline-ready

🎨 Modern UI — Clean, responsive design with Tailwind CSS

✨ Features 🤖 AI Assistant

Natural language campus queries

Context-aware responses (RAG)

Real-time streaming answers (SSE)

Multi-turn conversation support

LangGraph workflow with intent classification

🗺️ Interactive Campus Map

Leaflet-based real-time mapping

Dynamic route visualization

POI markers with clustering

Mobile-first & touch-optimized

🧭 Smart Navigation

Shortest-path routing (NetworkX)

Walking & accessibility modes

Turn-by-turn instructions

Distance & ETA estimates

Urgency modes (normal / exam rush)

📍 Campus Directory

Categorized facilities & POIs

Advanced search & filtering

Building, floor & room details

📚 Knowledge Management

Web scraping from official ASTU sources

Vector embeddings for semantic search

Document tagging & versioning

Admin content management interface

📱 Progressive Web App (PWA)

Installable on mobile & desktop

Offline support

Service worker caching

Push notification ready

🛠️ Tech Stack Frontend

Framework: Next.js 15 (App Router)

Language: TypeScript 5.8

UI: React 18.3

Styling: Tailwind CSS 3.4

State: Zustand 5.0

Maps: Leaflet + React-Leaflet

AI: Google Gemini

PWA: next-pwa

Backend

Framework: FastAPI

Language: Python 3.10+

Database: PostgreSQL (Supabase)

Vector Store: pgvector

Cache: Redis (optional)

AI Framework: LangChain + LangGraph

Routing: OSMnx + NetworkX

Scraping: BeautifulSoup + Trafilatura

🏗️ Architecture CLIENT (Next.js PWA) ├─ Hero Search ├─ Interactive Map (Leaflet) ├─ Campus Directory └─ AI Assistant (Chat UI) │ ▼ REST API / SSE Streaming │ ▼ SERVER (FastAPI) ├─ LangGraph AI Workflow │ ├─ Intent Classification │ ├─ Route Detection │ ├─ Context Retrieval (RAG) │ └─ Response Generation │ ├─ OSM Routing Service ├─ RAG & Vector Search └─ Web Scraper │ ▼ Supabase PostgreSQL + pgvector ├─ POIs ├─ Documents └─ Embeddings (768-dim)

🚀 Getting Started Prerequisites

Node.js 18+

Python 3.10+

PostgreSQL / Supabase

Google Gemini API Key

Docker (recommended)

🔹 Frontend Setup cd client npm install --legacy-peer-deps cp .env.example .env.local npm run dev

📍 Runs at: http://localhost:3000

🔹 Backend Setup cd server python -m pip install -r requirements.txt cp .env.example .env python migrate_db.py uvicorn main:app --reload

📍 API: http://localhost:8000

🐳 Docker Setup (Recommended) docker-compose up -d docker-compose logs -f docker-compose down

Services

Frontend → localhost:3000

Backend → localhost:8000

PostgreSQL → 5432

Redis → 6379

📡 API Documentation AI Query

POST /api/ai/query

{ "query": "How do I get to the library?", "latitude": 8.5523, "longitude": 39.2784 }

Navigation

POST /api/route

{ "origin": "Main Gate", "destination": "Engineering Block", "mode": "walking" }

Interactive Docs

Swagger → /api/docs

ReDoc → /api/redoc

📁 Project Structure divas/ ├── client/ # Next.js Frontend └── server/ # FastAPI Backend

(Detailed structure preserved exactly as your original — clean already 👍)

🔐 Environment Variables Client NEXT_PUBLIC_GEMINI_API_KEY= NEXT_PUBLIC_API_URL=http://localhost:8000

Server GEMINI_API_KEY= DATABASE_URL= SUPABASE_URL= SUPABASE_KEY= REDIS_URL=

🤝 Contributing

Fork the repo

Create a feature branch

Commit your changes

Open a Pull Request

✔ Follow best practices ✔ Write clean commits ✔ Update docs

📝 License

Licensed under the MIT License.

👥 Team & Contact

Developed by: 🎓 Computer Science & Engineering Club (CSEC), ASTU

GitHub: @CSEC-ASTU

Repository: https://github.com/CSEC-ASTU/divas

Issues: GitHub Issues

🙏 Acknowledgments

Next.js

FastAPI

Google Gemini

Leaflet & OpenStreetMap

Supabase

Tailwind CSS

Made with ❤️ by Team Divas

⭐ Star the repo — it really helps!