Trace

Trace (trace.so) is a workflow orchestration platform that gives AI agents shared context across a company's systems and lets teams build visual processes that combine AI and human steps. The product automatically identifies which tasks can be automated by AI and which require human input, with the stated aim of letting non-technical teams build and run workflows without engineering involvement. Its marketed positioning is an "AI orchestration layer for modern teams" — an operational layer that models how tasks flow between departments and aligns AI agents with human judgment.

Central to the product is what the company calls a context engine: by integrating with internal systems, it analyses internal wiki content, activity logs and people data, then builds a unified index (described as a context/knowledge graph) that supplies context for each new workflow and informs routing between people and agents. The platform respects existing permissions, learns from past and ongoing work, maintains a full audit trail, and allows users to review, edit or reroute tasks. Site material illustrates governance features such as approval gates (for example routing contracts above a value threshold to finance), confidence thresholds that flag low-confidence agent output for human review, and SLA monitoring across runs. The company describes a longer-term ambition of building a "CompanyOS" — a cognitive layer that ingests internal signals and drives action across teams.

The product was presented as being in beta on the company website, with a free demo available at demo.trace.so and a launch post published through Y Combinator.

Founding story

Trace was founded in 2025 by Tim Cherkasov (CEO) and Artur Romanov (CTO), who met at university and had worked together on projects, papers and startups over the preceding decade. Cherkasov was previously a product manager at digital asset custodian Copper, where he delivered an institutional offering for State Street, with earlier fintech experience as a data scientist across hedge funds and scaleups, and had been building startups since age 17 in retail tech, travel and web3. Romanov was previously a senior engineer at Deliveroo, where he built a grocery data-enrichment ML pipeline covering millions of items and thousands of retailers, and is a multi-time founder who led backends in logistics and finance and deployed workflow automation across engineering and operations.

Business model

B2B SaaS: an enterprise software platform sold to companies to orchestrate workflows across AI agents and employees. Y Combinator categorises the company under SaaS, B2B, productivity, enterprise software and AI. No pricing or contract terms are disclosed in the available sources.

Not disclosed in the available sources; the company offers a free demo environment (demo.trace.so) and asks prospective users to get in touch.

Traction

The Y Combinator launch post states that more than 30 companies use Trace to automate repetitive work. Team size is listed as 3. Product is described as in beta. Metrics displayed on the website (task counts, entities mapped, departments connected) appear in illustrative product mockups rather than as company disclosures.

Latest developments

A $3 million seed round was reported in February 2026, led by Y Combinator with participation from Zeno Ventures, Transpose Platform Management, Goodwater Capital, Formosa Capital, WeFunder and angel investors Benjamin Bryant and Kevin Moore. The company said proceeds would go toward enhancing the AI workflow orchestration platform, deeper context-graph engineering, expanding product and engineering teams, and scaling adoption of AI agents in enterprise workflows.