Agentic-AI

Modular AI Chat Application with Multi-Modal Capabilities

About the Project

Agentic-AI is a comprehensive Flask application providing a modern web interface for interacting with Large Language Models (LLMs), performing regression analysis, and classifying images - all with real-time streaming capabilities and AI-powered explanations.

The application integrates multiple AI services and offers a responsive user interface designed for both text and image-based interactions. It's built with a modular architecture allowing for easy extension with new ML/AI services.

Features

  • LLM Chat Interface: Chat with large language models with Markdown rendering and real-time streaming responses
  • Regression Analysis: Query regression models directly or via natural language
  • Image Classification: Upload images for classification with detailed AI explanations
  • Streaming Support: Real-time streaming for both chat and image analysis with ability to stop generation
  • Home Lights Control: Control simulated home lights via chat or API using natural language commands
  • Mobile Responsive Design: Optimized for both desktop and mobile devices
  • Extensible Architecture: Easy to add new ML/AI services or UI features

Technologies Used

Python Flask JavaScript Ollama API HTML/CSS REST API Machine Learning

Architecture Overview

The application consists of several integrated components:

  • Flask Backend: Orchestrates API routing, service integration, and CORS
  • Ollama Integration: Connects to Ollama LLM API for chat and text generation
  • Regression Service: External ML service for regression predictions
  • Image Classification Service: External ML service for image analysis
  • Web Frontend: Modern UI with chat and image analysis features