Summary
Deepank Agrawal is a software engineer specializing in production-grade AI agents, with eight years of experience spanning full-stack development, low-latency C++ systems, and large-scale knowledge pipelines. At Atlas he built autonomous customer-support agents that raised response accuracy from 40% to 95%, architected RAG pipelines at scale, and cut inference costs by 40%, while also shipping internal agentic dev tooling from concept to production. He thrives on ambiguous, zero-to-one problems and moves fluidly between prompt engineering, infrastructure, and frontend/backend work. Earlier, he optimized high-frequency trading simulations in C++ (improving runtimes 30–35%) and has a strong competitive-programming pedigree from IIT Kharagpur. Outside work he switches gears on the badminton court or hunting for good coffee, and he’s comfortable leading projects that require both deep systems thinking and rapid product iterations.
8 years of coding experience
3 years of employment as a software developer
BTech + MTech Computer Science and Engineering, BTech + MTech Computer Science and Engineering at Indian Institute of Technology, Kharagpur
English