Welcome to my Agentic AI and Multimodal AI Agents repository! This repository showcases my work in building advanced AI agents with reasoning capabilities and the ability to process multiple modalities (text, images, audio, etc.). It includes concepts, models, and practical implementations of AI agents designed for various real-world applications.
-
Multimodal AI Agent for Web Search & Stock Analysis : This project demonstrates a multimodal AI agent setup that integrates different AI models and tools for web searching and stock analysis. The AI agents use the Groq model (powered by Llama-3.3-70b) to gather and analyze information from web search engines (DuckDuckGo) and financial data sources (Yahoo Finance) in a highly collaborative manner.
-
Multimodal AI Agent for News Summarization & Sentiment Analysis : Multi-modal AI agents have become a significant area of interest due to their ability to process and integrate information from multiple data sources. This project presents the design and implementation of a personalized news aggregator built using multi-modal AI techniques. The system collects news articles, generates summaries using language models, and performs sentiment analysis to deliver relevant and customized content to users. The agent leverages state-of-the-art AI models and frameworks, demonstrating how intelligent automation can be applied to enhance user experiences.
-
Multimodal AI Healthcare : This project leverages Multimodal AI Agents to analyze medical reports from various specialist perspectives, including cardiology, psychology, pulmonology, neurology, endocrinology, and immunology. Using LangChain, OpenAI's GPT models, and PyMuPDF, the system extracts insights from a PDF medical report, processes it through multiple AI-powered agents, and generates a comprehensive multidisciplinary diagnosis.
-
AI Fitness Coach : The AI-Powered Fitness Coach is a Python-based application that generates personalized workout plans using AI. It tailors fitness recommendations based on user inputs such as fitness level, goals, duration, and available equipment. This project eliminates the need for generic workout plans by leveraging OpenAI's language model to create dynamic, customized routines.