I design and ship multi-agent conversational systems — from intent architecture and secure authentication to grounded, production-ready voice experiences. Built an enterprise insurance voice agent on Google CX Agent Studio, backed by live REST APIs, Cloud Run, and BigQuery.
I'm a Computer Science engineering graduate from Chandigarh University with hands-on experience building an enterprise-grade insurance voice agent — integrating live REST APIs, Cloud Run, and BigQuery for real-time authentication and claims workflows. I'm a published IEEE researcher in NLP (Chunking Studio, 5th ICCCES 2026).
A multi-agent voice system built at EXL, plus independent projects spanning NLP, data analytics, and fairness auditing.
A secure, multi-agent conversational voice assistant for insurance policy, claims, and renewal journeys — built on Google CX Agent Studio with deterministic safety callbacks, signed-token authentication, and grounded responses from live backend APIs and BigQuery.
An agent-based customer support chatbot that routes queries through intent detection to specialized agents for order tracking, payments, refunds, and login support — with authentication for sensitive actions and session memory to keep multi-turn conversations context-aware.
An end-to-end analytics platform processing UPI transaction data through an ETL pipeline, with an Isolation Forest model for fraud anomaly detection and a Random Forest model predicting spending category — both served via a FastAPI backend and an interactive Streamlit dashboard.
A normalized SQL database and interactive Power BI dashboards for algorithm fairness analysis, using joins, subqueries, and aggregations to surface bias patterns.