Guild AI raises $44M to build an enterprise agent control plane — think GitHub but for AI agents

Summary

Guild AI raised $44M across two rounds closed within four months, a $30M Series A and a $14M seed round led by Google Ventures with participation from NFX, Coastal Ventures, Scribble Ventures, Accru Capital, and Web Investment Network.

The company builds a control plane for AI agent fleets deployed in enterprise infrastructure. CEO James Everingham, who worked on developer infrastructure at Meta, describes the core problem bluntly. Once multiple agents run in production, governance and observability become critical. "It's like gremlins. You know? The first one's fine until they start multiplying and taking over and pulling levers in your infrastructure."

Guild's platform provides centralized governance, access controls, audit trails, cost management, and risk analysis. One customer blew through an entire monthly inference budget in twelve hours without visibility. Another engineer didn't know they had overspent. Beyond cost, the platform measures what companies actually get from that spend—agent impact, not just consumption.

Where agents are deployed

Adoption is accelerating across company size and stage, though not evenly by industry. Tech-forward companies and those facing competitive pressure move fastest. Regulated businesses move slower by design, though the calculus is shifting. The risk of moving slower now outweighs the risk of deploying agents in infrastructure with proper controls.

The most innovative use cases surprised Guild. Customers built a codebase agent that converses about system architecture and feature locations. Another agent analyzes code diffs for risk, letting low-risk changes through CI/CD while blocking potentially destabilizing ones, effectively eliminating code freezes. A third agent prevents duplicate bug reports by checking incoming bugs against existing ones. None of these required direct engineering effort; they were configured via natural language.

Agent marketplace

Guild positions itself as vendor and model neutral to avoid lock-in as customers move between foundation models. The company is building an agent hub, similar to GitHub, where companies can publish internal agents publicly or distribute third-party agents. A customer or competitor building agentic products can distribute agents through the hub; customers authorize and deploy them into their workspace within minutes.

This sidesteps the white-label question. Guild operates the infrastructure layer regardless of whether agents are built internally, bought as a product, or sourced from a marketplace.