Evos Lands $9M to Deploy AI Agents That Learn From Experienced Workers

Evos, formally registered as Evolutionary Operating System Inc., announced a $9 million pre-seed funding round to accelerate AI-driven automation adoption among smaller enterprises. The round was co-led by Uncork Capital and Field Ventures, with participation from Crosslink Capital, Utopia Capital Management, Sandwith Ventures and several undisclosed angel backers.

The company has built an agentic operations platform designed to create autonomous workers capable of handling tasks in industries including manufacturing, freight and logistics. Rather than replacing human workers outright, Evos trains its AI agents by having them observe and learn from experienced employees before taking on work independently.

Founder and Chief Executive Urav Shah identified a critical challenge facing smaller industrial firms: the rapid retirement of seasoned workers whose specialized knowledge remains undocumented. "The most capable system in many of these businesses is the person who has been there for 20 years, and what they know has never been written down," Shah said. Most companies with fewer than 100 employees operate through deeply ingrained processes that exist primarily in the minds of their most experienced staff.

Learning on the job

Evos' proprietary platform, called Workhorse, captures this institutional knowledge by automating the learning process. Agents assigned to specific tasks—such as handling broker communications, tracking missing documents, reconciling invoices or assembling freight loads—begin by observing rather than acting.

Each agent operates behind an "approval gate" during its training phase, requiring human review and authorization before executing any decision or action. Correct decisions become training data that strengthens the underlying model. Only after demonstrating full competency in real-world conditions under continuous human supervision does an agent transition to autonomous operation. When encountering unfamiliar situations, the agent must still obtain human approval before proceeding.

"An Evos operator learns the job from these people, improves on it and proves it can match them before it takes on any work by itself," Shah emphasized. "Our goal is to put that expertise on every job, every day, so these businesses can run like the best in the world do."

Shah acknowledged limitations in the technology but highlighted its practical performance. Agents typically achieve approximately 82% decision accuracy after 24 hours of observation, enabling them to autonomously handle roughly 70% of a human worker's daily tasks. Performance continues improving beyond this initial threshold.

Field Ventures Partner Richard Kerby noted that most businesses in the physical economy rely on software primarily for documentation and tracking rather than execution. "Evos changes that," Kerby said. "Its autonomous operators learn the judgment behind how a business runs, work within existing systems and take ownership of business functions."

Freight logistics company MyFreightWorld deployed Evos agents to validate less-than-truckload carrier invoices against its transportation management system. During its initial two-month deployment window, the agents completed more than 3,500 auditing decisions with 97% accuracy while processing invoices totaling over $600,000.

"Evos has helped us revolutionize our billing process with zero upfront investment and do it at an impressively fast speed measured in weeks and days," said MyFreightWorld President Mike Chalfant.

Evos launched with select customers just one month prior to the funding announcement and has already surpassed $1 million in annual recurring revenue. The company intends to release Workhorse for general availability later this year and will deploy the new capital toward expanding the rollout and building out its workforce.

Uncork Capital's Tripp Jones highlighted the speed at which organizations can operationalize Evos' technology. "It puts powerful technology within reach of businesses that software has long overlooked," Jones said.