At Epsilon, I architected a Databricks-based LLM AI agent that reduced manual attribution and pipeline diagnostic time by 80%. I also engineered Python and SQL tools to validate data quality across 15+ pipelines.
At Five Angstrom LLP, I constructed an academic ERP supporting 10+ user roles and created a compound database platform handling 120M+ molecular records. I also built protein visualization tools that reduced lookup time by 35%.
My projects include a real-time navigation application with train tracking and emergency services, and a senior-citizen FD comparison platform with personalized recommendations. I’ve also worked on an IoT home automation system and received an Oracle Cloud Infrastructure Data Science Professional and AI Foundations Associate certification.
