Hussein380/AI-Mpesa-Accountant

# AI-Pesa Project AI-Pesa is an AI-powered financial assistant that helps users track and manage their M-Pesa transactions through an intuitive web interface with AI chat capabilities. ## Project Overview AI-Pesa allows users to: - Upload M-Pesa statements or paste M-Pesa SMS messages - Get AI-powered analysis of their transactions - View categorized spending and financial insights - Chat with an AI assistant about their finances - Receive personalized financial advice ## Project Structure ``` / ├── frontend/ # Next.js frontend application ├── backend/ # Express.js backend with Gemini AI integration ├── docs/ # Project documentation ``` ## Key Features ### 1. Standardized API Responses All API responses follow a consistent format: ```json // Success response { "success": true, "message": "Operation successful message", "data": { /* Response data */ } } // Error response { "success": false, "error": { "message": "Error message", "code": "ERROR_CODE" }, "statusCode": 400 } ``` See API Reference for detailed endpoint documentation. ### 2. Model Consistency To ensure data consistency between frontend and backend, we use a shared type system: - TypeScript interfaces in the frontend match MongoDB schemas in the backend - All models are defined in `frontend/src/types/models.ts` - Consistent field naming and types across the application See Model Consistency Guide for more details. ## Frontend The frontend is a modern, responsive web application built with Next.js that provides: - User authentication - Transaction upload interface - Financial dashboard - AI chat interface - Responsive design with dark mode support See the frontend README for more details. ## Backend The backend is an Express.js application that: - Processes M-Pesa transaction data - Integrates with Google's Gemini AI for analysis - Stores user data and transactions in MongoDB - Provides RESTful API endpoints for the frontend See the backend README for more details. ## Sys …