Built AI-driven automation workflows and data enrichment pipelines at Hemut, a YC-backed logistics startup. Designed 5+ agent-based systems and API integrations to automate internal processes and support GTM operations.

Built AI features, ML pipelines, and RAG systems at Deepiri, a research collective of 35+ developers from top universities developing autonomous AI systems and productivity platforms, supporting 100+ internal users and improving research efficiency by 40%.

Designed an end-to-end ML pipeline that improved emotion classification accuracy from 70% to 88% through systematic feature engineering. Deployed the system to Google Cloud Platform with optimized performance, serving 10+ active users.

Built production APIs that reduced transaction failures by 45% and handled thousands of concurrent requests. Developed an automated archival system that cut storage costs by $57,600 annually through intelligent data management.

Built an internal chatbot platform that increased research efficiency by 40% for 100+ employees. Implemented comprehensive testing infrastructure that improved code coverage from 0% to 85%, catching 12 bugs before production.