Sakana AI has introduced Aquarium, a platform designed to stop companies from rebuilding the same business knowledge every time they deploy an AI agent. Announced on October 23, 2026, Aquarium packages data connections, definitions, decision rules and operating controls into a shared foundation that multiple agents can reuse.
At the center is an ontology — an organized map of business terminology, data relationships and decision criteria. Aquarium connects to tables and documents while leaving source data in its existing storage locations, so updates flow through to every agent built on top of it.
Workflows on the platform can combine information retrieval, AI judgment and function execution, with human approvals and a full record of the information consulted, decisions made, approvals received and results. The company also promises visibility into execution speed, accuracy and model usage.
The platform draws on earlier finance projects with Mitsubishi UFJ Bank, SMBC Group and Daiwa Securities Group — experience, the company stresses, rather than new deployments. Adoption starts with customer-specific consultation, and Sakana has disclosed neither pricing nor a general-availability date.
Aquarium lands as agent infrastructure becomes the real product battleground, alongside open agent projects like OpenHands, Cline and Goose.
Source: Superpower Daily
