Chetan Satkar

Architected and deployed an Agentic AI Interviewer platform utilizing LangGraph, orchestrating multi-agent interactions with adaptive difficulty logic and LLM-driven answer scoring to deliver structured, high-quality feedback.

Developed a JEE-level Math Monitor ChatBot incorporating a Hybrid RAG pipeline (FAISS + HuggingFace embeddings) and integrated LLM APIs, significantly enhancing problem-solving accuracy and context-awareness.

Engineered LLM evaluation engines that generate structured JSON outputs, automating assessment and feedback processes, and substantially reducing manual review efforts.

Applied fine-tuning techniques (LoRA/QLORA/PEFT) on LLaMA models, significantly improving model relevance and response quality for domain-specific hiring tasks.

Containerized and deployed AI agent systems using Docker and Kubernetes, ensuring reliable, production-grade performance and scalability.

Developed an intelligent Event Booking Agentic AI System on LangChain, orchestrating autonomous agents to automate complex multi-step booking workflows via a Flask REST API, enhancing operational efficiency.

Designed and implemented a full-stack Patient-Doctor Management System using Django, ensuring secure medical record handling, robust user authentication, and a scalable MVC-style backend architecture.

Created and deployed REST APIs using Flask and FastAPI to expose critical medical data services, facilitating seamless integration with frontend and third-party platforms.

Integrated AWS services (Lambda, S3) for scalable storage and serverless workflows, achieving reliable and cost-efficient deployment of AI solutions.

Collaborated on model evaluation and data preprocessing pipelines using Python (Pandas, NumPy, Scikit-learn) to develop and support AI-driven healthcare features.

Performed in-depth data analysis on Hospital ER datasets using Python (Pandas, NumPy) and SQL, extracting actionable insights that directly informed healthcare decision-making.

Developed and rigorously evaluated ML models on real-world medical data, applying advanced preprocessing and feature engineering techniques with Scikit-learn to significantly enhance model performance.