ResearchMind: Multi-Agent AI Research System
Technologies: LangGraph, StateGraph, LangChain, Tavily, Hugging Face Llama 3, BeautifulSoup, Streamlit
Developed a multi-agent AI research assistant to autonomously search, analyze, synthesize, and critique research topics. Implemented specialized Search, Reader, Writer, and Critic agents powered by LangChain, Tavily, Hugging Face Llama 3, and BeautifulSoup for autonomous research orchestration. Built an interactive Streamlit dashboard to visualize real-time agent execution and generate structured, fact-checked research reports with AI-driven feedback.
Vehicle Insurance Prediction MLOps Pipeline
Technologies: MongoDB Atlas, AWS S3, Docker, GitHub Actions, AWS EC2, ECR
Developed an end-to-end MLOps pipeline covering data ingestion, validation, transformation, model training, evaluation, deployment, and prediction workflow. Integrated MongoDB Atlas, AWS S3, centralized logging, exception handling, and configuration management to build a production-ready machine learning workflow. Designed a CI/CD-ready deployment architecture using Docker, GitHub Actions, AWS EC2, and ECR for scalable machine learning applications.
Mental Health Signal: ML-Based Score Prediction System
Technologies: FastAPI, Regression Model, Cloud Application
Developed a full-stack machine learning application that predicts a user’s mental health score from lifestyle, academic, and digital habit features using a trained regression model. Designed an intuitive FastAPI-powered prediction service with an interactive web interface, enabling real-time inference through a deployed cloud application. Implemented end-to-end data preprocessing, feature engineering, model training, and deployment to deliver an accessible AI-powered mental health assessment tool.
Languages
AI/ML
Frameworks/Libraries
Databases
Tools & Platforms
Cloud & DevOps
Other Technical Skills