Sanket Patil

I’m an AI Engineer with 2 years of experience building LLM agents, RAG pipelines, autonomous workflows, and scalable AI systems. I focus on retrieval systems, automation, and real-time intelligent applications that perform reliably in production.

At Bajaj Finserv, I developed enterprise-scale conversational voice calling agents to automate outbound call center operations for Personal Loans, Gold Loans, and Insurance. I designed a real-time pipeline (Speech-to-Text → LLM → Text-to-Speech), integrated multiple Azure OpenAI models, and built voice automation systems that handle thousands of calls with high accuracy and minimal latency.

I’ve also shipped automation products using Copilot Studio and enterprise search workflows—like an Autonomous Email Agent that reduced manual replies by 70% and improved SLA compliance by 40%. For retrieval-heavy experiences, I built real-time semantic search with <1.5s latency using ChromaDB + GPT-4, and delivered multi-user live interactions with FastAPI and WebSockets.

My strengths span LLM engineering (LangChain, LangGraph, LlamaIndex, guardrails, function calling), deployment and MLOps (MLflow, Docker, Kubernetes, CI/CD with GitHub Actions/Azure Pipelines), and backend development with Python, FastAPI, and REST APIs. I enjoy taking ideas from prototypes to production-grade systems with measurable impact.