Himesh Pandey

I’m a backend developer with a strong foundation in AI and systems engineering, driven by a passion for building intelligent infrastructure at scale. I focus on engineering the backend layer that powers next-generation AI applications—from core APIs to multi-agent AI systems.

In my Financial Technology internship at Uparjan Enterprises, I optimized 5+ financial service workflows, improving task completion rates by 20% through redesigned user journeys. I also enhanced backend APIs and data handling, cutting average processing time by 15%, and translated requirements into structured technical specs to deliver features 2 weeks ahead of schedule.

My projects reflect how I ship production-grade systems: I architected and deployed an MCP Artifact Store to reduce inter-agent context payload from ~1.6KB to 12 bytes per handoff, with atomic audit logging and per-artifact access control. I also built a RAG Pipeline Optimizer that reduced retrieval latency by ~40% and improved answer quality consistency to 90%+, while lowering inference costs by 35% and cutting hallucinations by 60% through guardrails.