Kanu AI exits stealth with $11.7M to turn employee workflows into owned software

Seattle-based Kanu AI said Wednesday it raised $11.7 million and emerged from stealth with a platform that watches how employees work and compiles those workflows into production software the company owns inside its own cloud.

Enterprises keep renting generic AI tools that never learn how their own teams actually work. Kanu’s pitch is that the judgment stuck in inboxes and spreadsheets becomes software the company owns and keeps, instead of another rented login that forgets the work tomorrow.

On Wednesday, 30 September 2026, Kanu AI announced $11.7 million in funding and said it was coming out of stealth. Stealth, here, means the company had been building without a public product launch. Trilogy Equity Partners led the funding. a16z speedrun, BMW i Ventures, and Accel took part. The wire calls it a round. It does not print a valuation. PR Newswire carries the release at 10:00 a.m. Eastern. The dateline is Seattle. The source line is Kanu AI. Those lines are Kanu’s.

The headline on that release is “Kanu Emerges from Stealth to Build AGI for the Enterprise.” AGI is short for artificial general intelligence, a phrase for software that could take on a wide range of work rather than one narrow chore. Kanu uses the phrase as the aim in the headline. The product the release actually describes is software built from a company’s own workflows. The release does not print a test, a score, or a date for a general system.

Kanu says it was founded just over a year ago. Karan Grover is founder and chief executive. The company says revenue has more than doubled in each quarter since launch. More than doubled means the next quarter’s revenue was more than twice the quarter before, on Kanu’s own books. It says it serves enterprises in commercial real estate, insurance, financial services, and technology. An enterprise, here, is a large company with its own security rules and its own software stack. Those lines are Kanu’s. The release does not name the customers in those industries.

The problem, in the release’s words. A company’s most valuable know-how rarely sits in a software product alone. It sits in how employees read the data, make the call, and get a result. That judgment is often stuck in people’s heads, in inboxes, in spreadsheets, and in one-off work. Employees repeat the work. The company’s systems do not learn it. The organization never gets to build on it. Build on it, here, means the next person starts from what the last person already figured out. Those lines are Kanu’s.

What the product does, as Kanu describes it. Staff show Kanu how they work and connect the systems they already use. Kanu then turns those workflows into operational software. A workflow is the sequence of steps a person already follows to finish a job. The software runs in the customer’s own cloud and inside the customer’s security layer. A cloud, here, is the bank of computers the company already rents, such as an Amazon Web Services account or a Google Cloud account. Data stays with the customer, reached through the permissions and policies the company already set. The customer keeps the choice of model, the logic, and the outputs. A model, here, is the AI system that reads the files and drafts the work. Those lines are Kanu’s.

Karan Grover, founder and chief executive, said: “Companies have spent decades buying software built around someone else’s idea of how they should work.” He said that instead of renting generic intelligence from another vendor, companies can use Kanu to turn their own data, judgment, and operating expertise into an asset they own, “and compound it across the organization.” Compound, in that sentence, means the next team starts from work the company already captured. That quotation is his, in the release.

For one national customer, Kanu says it cut an analysis that took up to eight weeks down to under 10 minutes. Eight weeks is about two months on the calendar. Ten minutes is one sitting. The work, Kanu says, ran across PDFs, email, spreadsheets, CRM data, and GIS systems. CRM is the system that holds customer records. GIS is mapping software for places and boundaries. Kanu says that customer is on track to cut software costs by more than $1 million this year and to bring in millions more in revenue by the end of the year. The release does not name the customer. It does not print the revenue figure past the word millions. It does not say an outside firm timed the work or checked the dollars. Those lines are Kanu’s account of its own deployment.

Four points the release lists for how a large company would actually run it. The software stays inside the customer’s cloud and security layer. When a workflow changes, Kanu says the software changes with it: one person teaches Kanu something, and the rest of the organization can use that the same day, without a rebuild or a long rollout. Every output can be opened. A person can see what Kanu read, which information it used, how it read that information, and why it reached its conclusion. Where a person still has to judge the work, the workflow can include a human review and an approval stop. The about box says every result can be inspected and every step that matters stays under human control. Those lines are Kanu’s.

Chuck Stonecipher, managing director of Trilogy Equity Partners, said the top of every enterprise leader’s list is how to use AI “to truly enhance their business performance in an enterprise-worthy fashion.” Enhance, here, means improve. Enterprise-worthy means fit for a large company’s security and approval rules. He said Kanu’s solution “uniquely answers that question.” Josh Lu, general manager and partner at a16z speedrun, said the firm believes in a future “where code is a commodity but doesn’t feel like a commodity.” A commodity, here, is something cheap and interchangeable. He said Kanu makes generated workflows “truly production-grade: stable, scalable and secure.” Production-grade means fit to run on real work. Scalable means it can take more of that work. Kasper Sage, managing partner at BMW i Ventures, called Kanu “a foundational technology for enterprise software” and said the team is giving organizations a way to “safely and reliably ship workflows without relying on long, manual engineering cycles.” Those quotations are theirs, in the release.

Kanu says the product is available today on AWS Marketplace and on Google Cloud Marketplace. AWS is Amazon Web Services. A marketplace, here, is the store inside that cloud where a company can subscribe to software. Kanu says it is hiring across engineering and business roles. The about box says Kanu turns the way an organization works into software workflows that organization owns, deployed inside the customer’s own cloud. Headquarters are in Seattle, Washington. The public site is getkanu.com. The media contact on the release is Mary Magnani at CodePR, marym@codepr.com. The release does not print a price and does not print a headcount. Those lines are Kanu’s.

The company homepage, the same day, carries a banner that reads “Read our $11.7M Seed announcement.” Seed, on that banner, is the company’s label for this funding. The wire itself says funding and round. The page says Kanu deploys inside the customer’s AWS or Google Cloud environment, and that the customer keeps the model accounts and the cloud bill. It prints four figures of its own: 5 times faster workflow completion, 74 percent weekly active adoption, 80 percent of outputs accepted without rework, and 24 hours saved per team per week. Weekly active adoption is the page’s label for people coming back during a week. The page does not say how it counted them. The page does not say which customer those figures come from. It does not tie them to the eight-week analysis or to the $1 million software-cost line. A navigation list names commercial real estate, private equity, financial services, and insurance. Private equity is on that list. The release’s sentence about who Kanu serves names commercial real estate, insurance, financial services, and technology, and does not use the words private equity. Those lines are on getkanu.com.

GeekWire’s John Cook reported the announcement the same morning. His account matches the investors, the revenue line, the industries in the release, the two marketplaces, and Grover’s quotation. On the customer example, GeekWire says the process went from eight weeks to under 10 minutes, and that software costs are expected to fall by over $1 million. The release’s wording is up to eight weeks and more than $1 million, and it also says millions in additional revenue. GeekWire adds that Kanu employs 10 people and plans to reach about 15 by the end of the year, and that the company plans to use the capital to hire. It says Grover was a machine-learning engineer and a three-time founder, a Y Combinator alumnus who previously worked as an engineer at Amazon and at DoorDash, and that an earlier startup of his built 3D audio tools so blind people could hear a room and move through it. Y Combinator is a program that funds very young companies. The biography and the headcount are GeekWire’s. They are not in the press release.

In plain terms, Kanu said on Wednesday that it raised $11.7 million, led by Trilogy Equity Partners, and came out of stealth with software that turns the way a team already works into programs the customer runs in its own cloud. The eight-week example and the $1 million software-cost line are the company’s account of one unnamed national customer. The homepage’s seed label, and its 5-times and 74 percent figures, are separate lines on getkanu.com.

The picture is Karan Grover on a stage. He wears a light gray sweatshirt with Kanu AI across the chest and holds a microphone. Behind him, the screen shows the AWS logo and the Google Cloud logo. It is a photograph of the founder at a cloud-partner event, issued with the funding announcement. The frame does not show the product’s screens, and it does not print a date.