gahana gupta

I’m a Full Stack Developer & AI Engineer with 3+ years building production LLM systems and scalable backend microservices. I focus on shipping reliable AI experiences that perform in real environments—not just demos.

In my recent role, I architected an AI Interview Assistant end-to-end using OpenAI & Gemini, delivering real-time LLM evaluation that’s live in production and serving thousands of users. I built RAG pipelines over internal datasets with vector stores and created an automated eval framework (hallucination rate, precision & recall) to keep prompt quality measurable.

I also optimized LLM spend by reducing API costs ~25% using Redis response caching, token-budget management, and output filtering. To handle high-concurrency AI workloads, I designed async microservices (Node.js + Redis queues), implemented streaming APIs with retry/fallback, and added guardrails for safer responses.

Before that, at AML RightSource, I designed and scaled KYC identity-verification microservices handling 10,000+ daily transactions with 99.9% uptime and zero data loss. I integrated real-time document verification pipelines that reduced manual review by 20%, built an API Gateway with JWT auth & RBAC, and led a monolith-to-microservices migration deployed on AWS + Docker with autoscaling and monitoring via Grafana & New Relic.