Gimlet Labs builds a multi-silicon cloud for AI inference, disaggregating workloads across heterogeneous accelerators to deliver faster tokens with greater power efficiency. Gimlet works at the intersection of AI research, production systems, and hardware, with a team based together in San Francisco. Its founding team includes Zain Asgar, Michelle Nguyen, Omid Azizi, Natalie Serrino, and James Bartlett.