Sajeev Magesh and Rohan Vij are the Co-Founders of ReasonBlocks, a YC-backed AI runtime layer that makes AI agents cheaper and more reliable by correcting reasoning and compressing tokens mid-run.

Sajeev studied Computer Science at Stanford and published research in Nature Sustainable Agriculture at the intersection of ML and agriculture, while Rohan studied Information Systems and AI at CMU and conducted molecular dynamics and AI research.

Sajeev is a USAMO qualifier and UN Best Paper award winner, bringing elite academic credentials to the AI infrastructure space. Rohan brings hands-on industry experience in AI for distributed energy at ENGIE and IoT research at UC Davis.

The duo has been building together for 11 years, a rare founding partnership that traces back to childhood, giving ReasonBlocks an exceptionally cohesive and battle-tested team dynamic.

ReasonBlocks has demonstrated a 52% token reduction and 42% accuracy lift on SWE-Bench Pro, backed by Y Combinator in the Spring 2026 batch with $500K raised.