Leena is an AI-powered sales assistant designed to support informal market vendors in resale purchasing decisions. The focus is to enable fast, assertive decisions that reduce waste and increase profit margins.
- Automatically analyzes the user's business profile and current stock without explicit input.
- Incorporates real-time external data: weather (via wttr.in) and upcoming public holidays (via Nager.Date).
- Generates personalized recommendations about what to buy, which items to avoid, and suggested quantities.
- Interacts with users via a simple fixed WhatsApp menu.
Note: Currently Leena analyzes only the user profile and current stock. Sales history analysis will be implemented in a future release.
- Street vendors and stall sellers
- Grocery shops and mini-markets
- Small neighborhood merchants without ERP systems
- Entrepreneurs needing help with stock and pricing
Users interact with Leena through a fixed WhatsApp menu:
1. View sales tips
2. Update product stock
3. View my profile
4. Update my profile
5. Ask a question
6. Help
7. Exit
Sales tips (option 1) are generated automatically by the AI based on the data stored for the user — no need to describe the situation, just select the option.
- The user opens the menu and selects an option (e.g.,
1. View sales tips). - Leena retrieves the user's profile and stock from the database.
- Those data are combined with external signals (current weather and upcoming holidays).
- The AI processes the inputs and generates a personalized recommendation.
- Leena replies with concise suggestions and risk alerts directly via WhatsApp.
- Reduces losses from idle stock and perishable products.
- Increases stock turnover by aligning restocking with expected demand.
- Reduces impulsive or excessive purchases.
- Enables advance planning using holidays, weather, and regional seasonality.
- Simple menu-driven interface suitable for non-technical users.
| Layer | Technology |
|---|---|
| Backend | Node.js + Express |
| Baileys | |
| Database | PostgreSQL (hosted on NeonDB) |
| LLMs | Gemini-2.0-flash-preview · Llama-3.3-70b (Groq) |
| Weather API | wttr.in |
| Holidays API | Nager.Date |
| Backend Deploy | Render |
| Landing Page | HTML + CSS + JS (Vercel) |
| Version Control | Git + GitHub |
This project is licensed under the MIT License.