At Netralabs, I engineered a Python pipeline that uses Tesseract OCR and Gemma to extract information from non-standard PDFs into validated JSON schemas. I also developed an emissions-tracking prototype and deployed automated enterprise agent workflows with human-in-the-loop approval.
At CirrusLabs, I developed Python services on AWS Lambda and built ETL pipelines feeding Snowflake and MongoDB for AWS QuickSight reporting. Earlier, I developed Python REST APIs and optimized database queries.

