Passionate AI & ML Engineer skilled in developing, training, and deploying intelligent systems.

Umang Sharma  • Peerlist

I Architected an intelligent Multi-Agent AI system using LangChain and LangGraph with four specializedcollaborative agents — Research, Reader, Writer, and Critic — orchestrated via a structured pipeline toautonomously perform deep web research, content synthesis, and quality evaluation.• Integrated Tavily Search API for real-time web retrieval and OpenAI LLM for advanced reasoning, enabling thesystem to produce accurate, context-aware research reports significantly superior to single-agent chatbotapproaches.• Deployed a modular, scalable application via Streamlit with a clean interactive UI, demonstrating practicalimplementation of Agentic AI architecture, LLM orchestration, and AI workflow automation

ResearchMind

Python | Machine Learning | Flask | AWS | Docker •Built an end-to-end Machine Learning pipeline to predict vehicle insurance outcomes using structured customer data. •Implemented data ingestion, data validation, preprocessing, feature engineering, and model training modules following MLOps best practices. •Trained and evaluated multiple ML models and selected the best-performing model based on evaluation metrics. •Developed a Flask-based web application for real-time predictions. •Containerized the application using Docker and deployed it on AWS for scalable access. •Designed modular, production-ready code with logging, exception handling, and configuration management