Summary
Sandesh Ghimire is a software engineer and Amherst senior studying Computer Science, Mathematics, and Physics, focused on building high-performance AI systems and low-latency architectures. Currently at a stealth AI startup he develops agentic AI infrastructure, multi-agent orchestration, and LLM tooling, having previously shipped offline NLP and on-device model work. His background spans systems research (published at IEEE IISWC), simulation frameworks for compute-in-memory, and production Flutter/Firebase apps, reflecting a blend of research rigor and product delivery. He gravitates toward distributed systems and ML-at-scale problems and has practical experience compressing models and building retrieval pipelines. A former Break Through Tech AI fellow and IPhO national gold medalist, he pairs strong theoretical training with hands-on engineering in fast-moving startups. Based in New York, he brings a rare mix of hardware-aware systems thinking and full-stack AI infrastructure experience.
13 years of coding experience