At Scageon Technologies, I developed a fully local RAG application for private, secure document-based question answering. I built Python ingestion and retrieval pipelines and containerized the application for local deployment.
At Cognizant Technology Solutions, I developed a Python framework to automate ML metric selection and processing for financial datasets. I built pipelines for monitoring model performance and added data-quality and threshold-based checks.
At Tata Consultancy Services, I optimized SQL Server and BigQuery queries on 100M+ records, improving execution speed by 35%. I also developed an invoice recognition and validation pipeline using OCR, NER, and GCP services, improving invoice validation accuracy by 93%.
Earlier, I built ETL pipelines with Python, PySpark, and Google Dataflow, and developed a speech-to-text and sentiment analysis pipeline on GCP. My academic research project used a Python NARMAX model to analyze wetland and climate data.

