At VMock, I build production AI systems for resume editing, career coaching, and resume-building workflows. I developed a Google ADK multi-agent resume editor that routes work across seven specialist agents and built a LangGraph coaching backend orchestrating 16 specialized agents with resumable, human-in-the-loop conversations.
I've reduced manual effort by approximately 75% and processing time by approximately 90% across 50+ configuration workflows using LangGraph and FastAPI pipelines. I also designed Resume Builder REST services and microservices, including resume parsing that normalizes uploads into structured JSON at 90%+ field accuracy.
I ship end-to-end product experiences across React, Redux, FastAPI, WebSockets, MySQL, PostgreSQL, Elasticsearch, and AWS. My work includes a 10+ language Resume Builder UI, cover-letter generation, and a Bootstrap migration that reduced page load by approximately 20%.
Previously at Zepto, I cut rider fraud by 20% through face-verification integration and improved delivery-slot efficiency by 10% with payment-tied scheduling and Golang synchronization jobs. I also build reliable distributed systems, including a fault-tolerant Kafka and Redis task queue with CI/CD, observability, retries, and dead-letter routing.
