At L&T Technology Services, I build and deploy LLM-based applications, autonomous agents, and retrieval-augmented generation pipelines for production use.
I've developed agent-evaluation tools that assess ground-truth outputs, hallucinations, reasoning, and control flow. I also built plan-execute-validation agents, deterministic workflows, and autonomous systems for tool creation and dependency management.
My work includes autonomous data science, AI-powered log and vulnerability analysis, multimodal chat systems, and token optimization frameworks. I use Python, LangChain, LangGraph, prompt engineering, model evaluation, and open-source LLMs to create scalable AI automation solutions.
