Anthropic has introduced Claude Managed Agents — a new platform designed to handle the full infrastructure required to run AI agents.

Developers no longer need to build complex architectures from scratch. Instead, they can describe tasks in plain language or via YAML, while the system automatically provisions environments, connects tools, and manages sessions and scaling.

Development time is reduced by up to 10x — from months to days or even hours.

Early use cases highlight how quickly solutions can scale:
— Notion is deploying dozens of agents within workflows;
— Asana is creating “AI employees” for task execution;
— Sentry is automating code fixes;
— Rakuten integrated agents across sales, marketing, and HR teams in just one week.

The launch signals a strategic shift for Anthropic — from competing at the model level to offering a full-stack platform combining models, infrastructure, and orchestration.

This move could reduce demand for standalone startups focused solely on “agent infrastructure,” as integrated platforms begin to dominate the space.