Kalavai, founded by Cambridge Judge alumni, unveiled an open‑source AI infrastructure platform aimed at turning underutilized GPUs into a distributed AI cloud — a pitch that resonates with university research computing units facing budget constraints for GPU time. Founders say their approach can cut inference costs up to 80% and unlock campus compute for faculty and student projects. Meanwhile, SAIL from Cornell Tech launched an enterprise AI tool for trade compliance, claiming pilot customers among Fortune 500 firms. The startup’s early adoption demonstrates how campus‑based AI projects are moving into regulated enterprise domains, where accuracy and auditability are critical. For university leaders, these companies flag opportunities and risks: partnerships can lower campus compute costs and create paid pilots for students, but commercial offerings also raise IP governance and data‑security questions for research collaborations.