ChatLomhat

A intelligent lomhat (teacher) assistant powered by AI that helps students with their studies and solves their academic problems. Designed to provide personalized learning support and educational guidance in a conversational format. Click here to visit ChatLomhat.

GenAIEducationPrompt

Project Overview

Purpose

ChatLomhat is designed to revolutionize the way students access educational support. Acting as an AI-powered teacher assistant, it provides personalized help with academic problems, explanations of complex topics, and study guidance tailored to each student's learning pace and style.

Key Features

Khmer language Support: Allows learners to study and interact in Khmer, making education more accessible and comfortable for native speakers.
Math Solving: Helps students solve math problems accurately, covering different levels and topics.
Step-by-Step Guidance: Provides detailed explanations in a structured way, so students understand the process, not just the answer.
Contextual Understanding: Understands questions within their context, ensuring more accurate, relevant, and personalized responses.
24/7 Availability: Always accessible anytime, anywhere, so learners can get support whenever they need it.

Technology Stack

AI/ML

Gemini Model

Flask-1.5

Backend

Python

FastAPI

Gemini API

Frontend

React

TypeScript

Tailwind CSS

Deployment

Vercel

Render

GitHub

Development Process

Environment Setup

Defines necessary environment variables such as the Google AI API key, Flask settings, and the frontend URL.

Dependency Management

Uses a requirements.txt file to manage Python dependencies.
Utilizes package managers like npm or yarn for frontend dependencies.

Web Framework Initialization

Sets up the Flask application with necessary configurations.
Initializes the React application with required dependencies.
Configures Cross-Origin Resource Sharing to allow requests from your frontend domain or localhost during development.

Third-Party Service Integration

Integrates with the Google AI API for advanced language processing capabilities.
Uses Pillow (PIL) to handle image input/output, which could be useful for exercise images or user-uploaded assets.

API Key and Configuration Management

Loads the API key from environment variables, not hardcoded in source, and raises errors if not found, enforcing good security practices.

Response Formatting

Cleans and formats AI responses to improve readability, removes code block markers, bullet points, and ensures proper step-by-step presentation for users.

Prompt Building and Request Handling

Builds prompts for AI requests, tailored for exercises, topics, difficulty levels, and number of problems.