Sakana AI announces 'Sakana Aquarium', an AI agent platform supporting decision-making and execution based on ontology

Sakana AI has announced 'Sakana Aquarium (hereinafter, Aquarium)', an AI agent construction, execution, and operation platform that supports an organization's decision-making and its execution based on ontology. Aquarium is a platform for efficiently constructing AI agents tailored to an organization and continuing to operate them safely. It organizes scattered information into common definitions using ontology, allowing AI agents to handle everything from information confirmation to judgment and execution.

The core consists of three mechanisms: Ontology, which systematizes the knowledge, rules, and data supporting organizational judgment; Governance, which controls data, ontology, and agents while recording judgment bases and execution results; and Managed Agents, which execute tasks based on ontology information and manage the entire agent lifecycle via described workflows.

Because prepared data and ontology can be reused by other agents, the company says it becomes easier to expand AI utilization within an organization. Aquarium aims to make accumulated organizational information usable for daily judgment and action, enabling rapid response to situational changes and agile execution under human involvement.