At Capgemini, I delivered an enterprise Generative AI-powered BI Copilot for a US utility client, extending its Power BI and Actuate BIRT reporting environment. I engineered a natural-language-to-SQL pipeline that reduced insight generation time from 2–4 hours to 5–10 minutes.
I designed a RAG knowledge layer with FAISS and OpenAI embeddings to retrieve KPI definitions, business rules, data dictionaries, and report metadata. I also added SQL validation and read-only safeguards for queries against sensitive data.
At NTT Data, I led the migration of legacy BRIO reports to OpenText Magellan BI with zero data loss, and designed data objects and custom reports. At Mphasis, I served as Scrum Master and lead report developer, improving team velocity by 25% through sprint process improvements.
Earlier, I built BI reports and dashboards for finance, healthcare, insurance, and banking clients, and developed full-stack web applications using ReactJS, Node.js, and MongoDB. My project work includes a RAG document Q&A system, a BERT sentiment classifier, and a churn prediction pipeline.
