Sail Research - Cost-efficient AI inference API for long-horizon agents.
SF Compute - Sublease supercomputer GPU clusters and resell unused capacity.
