At VERTIV ENERGY, I developed enterprise applications that improved workflow efficiency by about 30% through automation and optimized processes. I also integrated REST APIs and enterprise services and designed data validation workflows.
I worked on Agentic AI initiatives using Python, LangChain, and LangGraph, designing, building, and deploying multi-agent diagnostic systems. I integrated LLM-based solutions into enterprise workflows and implemented a FAISS-based RAG pipeline for diagnostic queries.
On the Agentic AI Diagnostic Assistant project, I designed a multi-agent architecture for automated diagnostics of UPS, Battery, and Thermal systems. I combined historical equipment logs with real-time sensor inputs and built tool-calling workflows for log extraction and BigQuery queries.
I standardized model outputs with structured JSON schemas and developed a multi-model evaluation pipeline using LLM-as-a-Judge scoring. At Nagarro, I developed POS, billing, and reporting systems used across 500+ business units, and improved performance by about 15% through optimized queries and data handling.
