Despite their vast capabilities, AI agents have been gradually penetrating the enterprise sector, and a new startup believes the slow progress is due to insufficient context.

Emerging from Y Combinator’s 2025 summer cohort, Trace is a workflow orchestration company set to address this void. The firm creates detailed maps of intricate corporate ecosystems and processes so that agents receive the necessary context for rapid scaling.

“OpenAI and Anthropic are developing these brilliant interns that can be utilized within the company,” says Trace CEO Tim Cherkasov, referring to the tools from the AI laboratories. “We’re constructing the manager that understands how to deploy them.”

On Thursday, the London-based startup announced it had secured $3 million in seed funding from Y Combinator, Zeno Ventures, Transpose Platform Management, Goodwater Capital, Formosa Capital, and WeFunder. Additionally, angel investors Benjamin Bryant and Kevin Moore contributed funds.

Trace’s platform begins by creating a knowledge graph from the company’s current tools—systems such as email, Slack, and Airtable that influence daily operations. With this context established, users can instruct the system with a broad task—such as “We need to create a new microsite” or “Let’s prepare our 2027 sales strategy”—and Trace will respond with a detailed workflow, assigning certain tasks to AI agents and others to human personnel. When activating an AI agent, the system provides it with the specific information required to complete its task.

The aim is to streamline the intricate process of integrating AI agents, which is one of the main obstacles to their actual implementation in enterprises.

Given the numerous companies concentrating on agent-based AI, Trace will face significant competition. Earlier this week, Anthropic unveiled its own version of enterprise agents focused on ready-made plugins for particular departmental roles. Additionally, numerous workplace productivity services that Trace will utilize, like Atlassian’s Jira, are introducing their own agents that could rival the startup’s platform.

However, the founders of Trace are confident that their knowledge-graph methodology will be crucial for achieving success, as they can embed context engineering deeply within the framework of agentic deployment.

“The years 2024 and 2025 were predominantly about prompt engineering. We’ve now transitioned from prompt engineering to context engineering,” states CTO Artur Romanov. “Whoever delivers the most effective context at the optimal moment will become the foundation upon which AI-first enterprises will thrive. We aspire to be that foundation.”