DigitalOcean, a cloud infrastructure provider focused on simplifying scalable computing for developers, announced the acquisition of Katanemo Labs, an agentic AI infrastructure company, to advance its Agentic Inference Cloud strategy. The deal expands DigitalOcean’s platform into the operational layer of AI agents, integrating orchestration, observability, and execution capabilities for production deployments. Katanemo Labs co-founder and CEO Salman Paracha joins DigitalOcean as Senior Vice President of AI, bringing expertise in model research and open-source infrastructure.
Bridging the Production Gap
Industry momentum has shifted from AI experimentation to production deployment, where reliability, safety, and observability remain key challenges. The acquisition addresses these gaps by adding core primitives such as orchestration, workflow execution, and governance layers designed to support scalable multi-agent systems. The integration aims to help organizations move AI projects from pilot stages into operational environments more efficiently.
Data Plane and Observability Innovation
Katanemo Labs contributes a framework-agnostic data plane developed through the Plano open-source project. The technology acts as a “NoOps” layer that abstracts operational complexity while enabling predictable performance. Signals-based observability capabilities provide insights into agent behavior, turning production traces into actionable intelligence for diagnosing failures and improving performance.
Expanding the Agentic Stack
Small Action Models from Katanemo Labs, including the Arch-Agent family, complement the data plane and provide flexible building blocks for autonomous systems. The combined platform offers AI teams integrated tools for deploying and scaling agents without fragmented infrastructure or additional operational overhead.