Bespoke Labs Raises $40M to Build Practice Worlds for Enterprise AI Agent Training

Bespoke Labs Raises $40M to Build Practice Worlds for Enterprise AI Agent Training

Bespoke Labs, an AI agent training platform that creates simulated environments for agents to practice in, has raised $40 million across a seed round led by 8VC and a Series A led by Wing. The company will use the capital to build out its enterprise agent training infrastructure.

The funding signals growing recognition that enterprise agent reliability — a key bottleneck in AI agent adoption — requires structured, environment-based training rather than simple prompt engineering. Bespoke's approach creates synthetic practice worlds where agents can fail safely and learn, addressing the critical gap between demo-level performance and production-grade reliability. This positions Bespoke alongside other emerging infrastructure plays that treat agent behavior as a training problem, not just a model capability problem.

While $40M is modest relative to the mega-rounds dominating foundation model headlines, the round fits a broader pattern of capital flowing into agent infrastructure — the "picks and shovels" of the agent economy. If Bespoke's environments prove they can materially reduce failure rates in enterprise deployments, the company could become a cornerstone of the agent tooling stack, though it faces competition from incumbents offering testing frameworks and observability platforms. The real test will be whether synthetic worlds generalize to the messy, unpredictable workflows of real enterprises.

#bespokelabs #agenttraining #simulation #enterpriseAI #seedfunding #seriesA