At TechPrysm, I designed LLM-backed semantic search that replaced legacy keyword search and reduced average query-to-result time from 4.6 seconds to 1.4 seconds. I also resolved Hibernate query performance issues, improving p95 API latency from 5.2 seconds to 1.3 seconds.
At LetsRemotify, I led a phased migration from a legacy J2EE/Struts monolith to Spring Boot microservices and transformed nightly batch reporting into near real-time pipelines. I’ve also built LLM-assisted features for learning and nursing platforms, and developed a cloud-based rental-management workflow.

