Balwant Singh Bisht

I’m a Data Scientist with both industry and research experience, focused on machine learning systems that work end-to-end—from data engineering to model behavior. At Zee Entertainment Enterprises Limited, I built a conversational business intelligence system that enabled natural-language querying by integrating Azure Speech-to-Text, GPT-4, SQL generation, and Google BigQuery.

I also built and maintained automated ETL pipelines using Python, Apache Airflow, REST APIs, and Google BigQuery to ingest and transform analytics data for downstream analysis. During an AWS-to-GCP migration, I validated data quality, updated Airflow workflows, and verified consistency through SQL-based testing, earning recognition as one of the top three contributors during a Microsoft–ZEEL hackathon. In parallel, my research as a Junior Research Fellow explored privacy-preserving, communication-efficient Federated Learning for secure AI in healthcare—where I designed a personalized Federated Learning framework combining client-specific pruning with Differential Privacy, evaluated trade-offs between accuracy, communication efficiency, and privacy, and contributed to a research manuscript.