Despite the immense potential of AI agents, their adoption in the enterprise sector has been gradual. A new startup, Trace, aims to address this challenge by providing the necessary context for effective implementation.
As part of Y Combinator's 2025 summer cohort, Trace is focused on workflow orchestration, mapping intricate corporate structures and processes to empower AI agents to scale effectively. CEO Tim Cherkasov emphasizes the need for a "manager" that understands how to integrate these intelligent tools into the workforce.
On Thursday, the London-based startup announced it has successfully raised $3 million in seed funding from a range of investors, including Y Combinator, Zeno Ventures, Transpose Platform Management, Goodwater Capital, Formosa Capital, and WeFunder. Notable angel investors Benjamin Bryant and Kevin Moore also contributed to the funding round.
Trace's innovative approach begins with constructing a knowledge graph from a company's existing tools--such as email, Slack, and Airtable--essential for daily operations. Once this context is established, users can input high-level tasks like "design a new microsite" or "develop our 2027 sales plan." Trace then provides a detailed workflow, designating tasks to both AI agents and human employees. When AI agents are involved, the system ensures they receive the precise data needed to execute their assigned tasks.
The goal is to streamline the onboarding process for AI agents, which has been a significant hurdle to their widespread adoption in businesses.
With numerous companies exploring agentic AI, Trace is entering a competitive landscape. Recently, Anthropic unveiled its own enterprise agents, focusing on pre-built plugins for specific business functions. Many productivity platforms that Trace relies on, like Atlassian's Jira, are also launching their own AI agents, presenting potential competition.
However, the founders of Trace are confident that their knowledge-graph methodology will set them apart, embedding context engineering into the very fabric of AI deployment.
CTO Arthur Romanov highlights the shift from prompt engineering to context engineering, asserting that the ability to deliver optimal context at the right moment will determine the success of AI-first enterprises. "We aspire to be the foundational infrastructure for these advancements," he states.