Munshi

demo.mp4

Aapke dukaan ka Munshi

Features • Tech Stack • Get Started • API

---

Munshi is an AI-powered mobile assistant designed specifically for small Indian local shops — kirana stores, chai stalls, cafes, and neighborhood vendors. It speaks to you in Hindi (Roman script), understands your voice, and helps you manage your entire shop from inventory to insights — without needing to be a tech expert.

Whether you want to know your top-selling products, check what stock is running low, or simply ask "Aaj kitna paisa gaya?" — Munshi has your back.

---

Features

Feature	Description
Voice & Text Chat	Talk to Munshi in Hindi — voice input gets Hindi responses spoken back to you
Stock Management	Add, update, and track products with purchase price, selling price, and expiry dates
Analytics Dashboard	Visual sales charts, revenue tracking, top/worst selling products
Transactions	Record sales with Cash, UPI, or Credit payment tracking
Quick Scan	Barcode scanning for instant product lookup
AI Conversations	Ask anything about your business — "Kis product se sabse zyada profit ho raha hai?"

---

Tech Stack

Server

• Runtime: Node.js + Express
• Database: PostgreSQL + Prisma ORM
• Auth: Better-Auth
• AI: OpenAI Agents SDK + Ollama (local LLM)
• Voice: Sarvam AI (Hindi STT/TTS)

Client

• Framework: Expo (React Native)
• Styling: NativeWind + Tailwind CSS
• Animations: Moti
• UI Components: Radix UI Primitives
• Icons: Phosphor Icons + Lucide

---

Getting Started

Prerequisites

• Node.js 18+
• PostgreSQL
• npm
• Ollama (for local AI)
• Sarvam AI API key (for voice features)

Clone & Install

# Clone the repository
git clone https://github.com/binit2-1/munshi.git
cd munshi

# Install server dependencies
cd server
npm install

# Install client dependencies
cd ../client
npm install

Environment Setup

Create .env files in both server/ and client/ directories:

Server .env:

DATABASE_URL="postgresql://user:password@localhost:5432/munshi"
BETTER_AUTH_SECRET="your-secret-key"
OLLAMA_BASE_URL="http://localhost:11434"
SARVAM_API_KEY="your-sarvam-key"
PORT=3000

Client .env:

EXPO_PUBLIC_BASE_URL="http://localhost:3000"

Run the App

# Terminal 1: Start the server
cd server
npm dev

# Terminal 2: Start the Expo app
cd client
npm dev

Database Setup

cd server
npx prisma migrate dev
npx prisma db seed  # Optional: seed sample data

---

API Overview

Authentication

Endpoint	Method	Description
/api/auth/signup	POST	Register new user
/api/auth/signin	POST	Login user
/api/auth/signout	POST	Logout user

Products

Endpoint	Method	Description
/products	GET	List all products
/products	POST	Create new product
/products/:id	PUT	Update product
/products/:id	DELETE	Delete product

Stock

Endpoint	Method	Description
/stock	GET	Get all stock levels
/stock	POST	Update stock quantity

Transactions

Endpoint	Method	Description
/transactions	GET	List transactions
/transactions	POST	Create new transaction

AI Assistant

Endpoint	Method	Description
/ai/response	GET	Text chat with AI
/ai/audio/response	POST	Voice chat with AI
/ai/history	GET	Get chat history

---

📱 Screenshots

---

How the AI Works

Munshi's AI agent uses the OpenAI Agents SDK with tools that execute database commands:

1. User asks in Hindi → "Mujhe batao konsa product kam becha?"
2. Agent parses intent → Calls appropriate CLI tool
3. Tool queries PostgreSQL → Returns JSON data
4. Agent responds in Hindi → "Aapka sabse kam bechne wala product Parle-G hai..."

All conversations are stored in chat history with Redis-powered sessions for context.

---

License

MIT License — feel free to use and modify for your shop!

---

Acknowledgments

• Sarvam AI — For Hindi voice capabilities
• Ollama — For local LLM inference
• Better-Auth — For seamless authentication

---

Made with ❤️ for local shop owners across India