For all their potential, AI agents have been slow to make a significant impact in the enterprise, and one new startup, Trace, is betting that the primary reason for this bottleneck is a critical lack of context. Launched as a standout member of Y Combinator’s highly anticipated 2025 summer cohort, Trace is a workflow orchestration startup headquartered in London, specifically designed to bridge this crucial gap. The company’s innovative approach focuses on meticulously mapping complex corporate environments and intricate business processes, thereby equipping AI agents with the deep, operational context they require to scale rapidly and perform effectively within the often-fragmented landscapes of large organizations.
On Thursday, Trace publicly announced it had successfully closed a seed funding round, securing $3 million from a diverse group of prominent investors. The round saw participation from Y Combinator, reinforcing its confidence in the startup’s vision, alongside Zeno Ventures, Transpose Platform Management, Goodwater Capital, Formosa Capital, and the crowdfunding platform WeFunder. Further bolstering this initial investment, notable angel investors Benjamin Bryant and Kevin Moore also contributed, signaling strong belief in Trace’s potential to redefine enterprise AI.
The Unfulfilled Promise of Enterprise AI Agents
The explosion of generative artificial intelligence and large language models (LLMs) since late 2022 has captivated the tech world and promised a new era of productivity. Businesses across every sector have eagerly explored how these powerful AI tools, particularly in the form of autonomous agents, could automate tasks, enhance decision-making, and streamline operations. However, the journey from proof-of-concept to widespread, impactful deployment within the enterprise has proven far more challenging than initially anticipated. Industry reports consistently highlight low adoption rates for complex AI systems beyond pilot projects, often citing integration difficulties, data privacy concerns, the potential for AI "hallucinations," and a general lack of trust as major impediments. Yet, a more fundamental issue persists: AI agents often operate in a vacuum, lacking the intricate understanding of specific company workflows, legacy systems, and unwritten rules that govern daily operations.
"OpenAI and Anthropic are building these brilliant interns that can be leveraged within the company," explains Trace CEO Tim Cherkasov, articulating the prevailing sentiment around the capabilities of leading AI labs’ foundational models. "We’re building the manager that knows where to put them." This analogy succinctly captures Trace’s strategic position in the AI ecosystem – not as another foundational model developer, but as the essential orchestrator that turns raw AI power into actionable, enterprise-grade solutions. The challenge is akin to hiring a highly skilled individual who, despite their talent, lacks any understanding of the company’s culture, processes, or even where the coffee machine is. Trace aims to provide that comprehensive operational blueprint for AI agents.
Trace’s Context Engineering: A Deep Dive into the Platform
At its core, Trace’s system is built upon a sophisticated methodology of "context engineering." The process commences by constructing a dynamic knowledge graph, meticulously piecing together information from a company’s existing digital infrastructure. This includes an array of ubiquitous enterprise tools – from communication platforms like Slack and email systems, to project management solutions such as Airtable, and potentially extending to CRM, ERP, and HR platforms. By integrating with these disparate data sources, Trace gains a holistic understanding of the firm’s day-to-day working life, identifying key entities, relationships, dependencies, and operational norms.
Once this comprehensive knowledge graph is established, users can interact with the system by issuing high-level, strategic prompts. These are not granular, step-by-step instructions but rather overarching objectives, such as "We need to design a new microsite" or "Let’s develop our 2027 sales plan." Upon receiving such a prompt, Trace’s intelligent orchestration engine springs into action. It leverages the detailed context embedded within its knowledge graph to decompose the high-level objective into a series of actionable, sequential steps. Crucially, it then intelligently delegates these sub-tasks, assigning some to human workers who possess unique insights or require final approval, and dispatching others to specialized AI agents.
When an AI agent is invoked, Trace doesn’t merely pass on the task. Instead, it provides the agent with a highly curated and specific subset of data drawn directly from the knowledge graph – the precise context needed to successfully complete its sub-task without needing to infer or search broadly. This targeted context injection is pivotal, minimizing the chances of irrelevant outputs, improving accuracy, and significantly reducing the "onboarding" time that typically plagues AI agent deployment. The idea is to automate away the delicate and often time-consuming work of hand-holding AI agents, which has historically been one of the biggest blockers for their actual deployment and scalability within companies.
Timeline and Strategic Positioning within the AI Evolution
Trace’s emergence during the Y Combinator 2025 summer cohort places it at a crucial juncture in the evolution of enterprise AI. The years 2023 and 2024 were largely characterized by intense experimentation with "prompt engineering" – the art and science of crafting effective inputs for generative AI models to elicit desired outputs. While prompt engineering remains relevant, the industry is rapidly maturing, shifting its focus towards more robust and scalable solutions.
Artur Romanov, Trace’s CTO, succinctly encapsulates this paradigm shift: "2024 and 2025 was still about prompt engineering. Now we’ve moved from prompt engineering to context engineering. Whoever provides the best context at the right time is going to be the infrastructure on top of which the AI-first companies will be built. And we hope to be that infrastructure." This statement underscores Trace’s ambition to become a foundational layer for future AI-driven enterprises, moving beyond mere instruction-giving to providing the foundational understanding necessary for AI agents to truly operate autonomously and intelligently within complex business processes. The timing of their funding and product launch aligns with an increasing industry demand for practical, deployable AI solutions that can deliver tangible ROI, rather than just impressive demonstrations.
Investor Confidence and Market Dynamics
The $3 million seed funding round is not just a financial milestone but a strong vote of confidence from a diverse group of investors. Y Combinator’s participation, known for its rigorous selection process and track record of identifying disruptive startups, lends significant credibility to Trace’s model. Investors in the current climate are increasingly looking beyond the foundational AI models themselves, seeking out "picks and shovels" companies that build the necessary infrastructure, tools, and platforms to make AI truly useful and scalable for businesses. Trace fits this profile perfectly, addressing a clear and pressing need in the market.
The global market for AI in enterprise applications is projected to grow exponentially, with some estimates reaching hundreds of billions of dollars within the next few years. However, this growth is contingent on overcoming the very integration and context challenges that Trace aims to solve. Investors like Zeno Ventures and Goodwater Capital, with their focus on early-stage technology companies, likely recognize the massive untapped potential in unlocking AI’s value in real-world business scenarios. Transpose Platform Management and Formosa Capital also signal a belief in the platform’s ability to create a new category or significantly disrupt existing workflow automation markets. The inclusion of WeFunder indicates a broader appeal and confidence from a wider investor base, including potentially those closer to the practical business applications.
Navigating a Crowded and Evolving Competitive Landscape
While Trace’s approach is distinct, the broader field of agentic AI and workflow automation is becoming increasingly competitive. Many companies are intensely focused on leveraging AI to streamline enterprise operations. Just earlier this week, Anthropic, a leading AI research company, launched its own initiative for enterprise agents, focusing on pre-built plugins tailored for specific departmental functions like finance, engineering, and design. This represents a direct competitive thrust, offering specialized AI capabilities within predefined contexts.
Furthermore, many incumbent workplace productivity services that Trace will integrate with are also developing their own AI agent capabilities. For instance, Atlassian’s Jira, a widely used project management tool, recently announced updates allowing AI agents and human workers to collaborate side-by-side within its platform. Similarly, other major players in CRM, ERP, and communication software are embedding AI functionalities directly into their offerings, potentially competing with Trace’s orchestration layer. Robotic Process Automation (RPA) vendors are also rapidly integrating generative AI to make their automation more intelligent and adaptive.
Despite this formidable competition, Trace’s founders remain confident that their knowledge-graph approach provides a critical competitive advantage. They believe that embedding "context engineering" deep into the very structure of AI agent deployment is a more robust and scalable solution than relying on generic plugins or siloed departmental agents. By providing a universal framework for understanding and orchestrating AI within any complex enterprise environment, Trace aims to offer a foundational layer that can complement, rather than be entirely superseded by, other specialized AI tools. The London base also positions Trace strategically within Europe’s rapidly expanding tech ecosystem, offering access to diverse talent and a market with unique regulatory and operational nuances.
Broader Implications and the Future of Work
Trace’s vision, if successful, carries significant implications for the future of work and enterprise productivity. By democratizing the deployment of AI agents and making them truly operational within existing business structures, Trace could accelerate digital transformation initiatives across industries. This shift would allow human workers to offload more mundane, repetitive, or data-intensive tasks to AI, freeing them to focus on higher-level strategic thinking, creativity, and complex problem-solving that still uniquely require human intelligence. The result could be a more efficient, agile, and innovative workforce.
However, the journey will not be without its challenges. Scaling Trace’s knowledge graph across enterprises of varying sizes and industries, each with unique legacy systems and data architectures, will be a complex undertaking. Ensuring robust data privacy and security measures, especially when integrating with core enterprise systems, will be paramount. Moreover, maintaining the accuracy and reliability of AI agent outputs – preventing "garbage in, garbage out" scenarios – will require continuous refinement of the knowledge graph and orchestration logic. Finally, user adoption and managing organizational change will be critical, as integrating AI agents fundamentally alters existing workflows and responsibilities.
As the AI landscape continues to mature, the focus is increasingly shifting from the raw power of models to their practical, reliable, and scalable application. Trace’s $3 million seed funding round marks a significant step in its mission to become the essential infrastructure for "AI-first" companies, positioning itself at the forefront of this critical transition from prompt engineering to context engineering. The company’s success could well define how enterprises finally harness the full potential of AI agents, transforming the promise of artificial intelligence into tangible business value.
About the Author
