Sakana AI launches Aquarium, an ontology-based platform for enterprise AI agents

Sakana AI announced Sakana Aquarium, a platform for building, running, and operating AI agents grounded in an organization's ontology. Its three components are Ontology, which structures knowledge, rules, and data without migrating them; Governance, which controls access, requires pre-execution approval, and records decision rationale and outcomes; and Managed Agents, which deploys workflows combining data reference, AI judgment, function execution, and human approval. Data and ontology can be reused across agents.

Key takeaways

• Ontology organizes knowledge, rules, and data with their meanings and relationships, connecting formats without migrating them from existing stores.
• Governance sets access permissions and pre-execution approvals and records decision basis through execution results.
• Managed Agents deploy workflows of data reference, AI judgment, function execution, and human approval, managing the agent lifecycle.
• Prepared data and ontology can be reused by other agents, allowing quick launch of operation-specific agents on one platform.

Why it matters

It targets a common enterprise problem: repeatedly rebuilding data connections and terminology for each AI deployment. A shared ontology and governance layer could make agent adoption more scalable and auditable inside organizations.

Read with care

Source claims not independently checked; the text is a product announcement with no reported performance metrics or customer outcomes beyond named partnerships.