A Bellevue startup that launched measuring AI spend just renamed itself around measuring AI value — and the shift tells you where enterprise AI actually is right now.
Pay-i emerged from stealth in May 2025 with a $4.9 million seed round. That round was aimed, in the company’s own description, at AI costmanagement. Enterprises were spending heavily on AI and had no clean way to track what they spent. Sixteen months later, the company no longer exists under that name.
On Wednesday, September 30, 2026, it announced a new name — Ascerta — and an $18 million Series A. Dell Technologies Capital led the round. Hitachi Ventures, BGV, Wipro Ventures, and earlier investors also participated. Total funding now stands at $22.9 million. No valuation was disclosed.
The rename is not cosmetic. The company’s own announcement opens with the thesis shift: enterprises can count tokens, licenses, agent runs, and lines of AI-generated code. Most still cannot answer which AI initiatives are actually worth scaling. That is the problem Ascerta now says it solves.
Ascerta was founded in 2024 by three Microsoft veterans. Chief executive David Tepper spent 19 years at Microsoft and led GenAI strategy for internal use across Azure. Cofounder Doron Holan spent 27 years there and architected hyperscale throttling infrastructure that the release says handled hundreds of billions of requests a day. The third cofounder is Erik Winters. The company is based in Bellevue, Washington.
Sourcing note: globenewswire.com does not serve this publication and ascerta.ai carries no funding post. The wire release was read as a mirrored full text. All figures and quotes below are drawn from that release.
The key insight: A company that started by counting AI spend renamed itself around AI value inside sixteen months. That is a legible signal. The spend question got answered. The worth-it question did not — and that is where the money is moving now.
These are the figures the company puts on its own product. They are printed here as claims, with what is missing stated alongside them, rather than as demonstrated results.
The Structural Read
The easy thing to meter is not the thing that decides anything.
Tokens, licenses, agent runs, and lines of generated code are all countable. That is precisely why they became the default metrics. And none of them tells a buyer whether to scale a project or kill it.
Tepper says it plainly in the release: the market is full of meaningless vanity metrics. Companies are counting tokens and agent runs. They are struggling to derive the actual impact AI has on their business.
This is a reading of one company’s repositioning, not a survey. One rebrand is an anecdote. It is still worth noticing when the anecdote comes from the people who were selling the first answer and found they needed to sell a different one.
Raman Khanna — Dell Technologies Capital
“Ascerta is building the system of record for AI value creation.”
A system of record is a strong claim for a category that did not have a name eighteen months ago. It implies permanence, stickiness, and the kind of data gravity that compounds over time.
The harder problem sits underneath the positioning. Measuring AI value means attributing a business outcome to a specific model call. That is the same attribution problem marketing analytics spent twenty years working on and never fully solved. Ascerta says its platform ties each use case to the KPIs it was meant to move. That is a claim about solving attribution. It is unverified.
FDE Framework Lens
Ascerta as Enabler — Not Builder, Not Distributor
Ascerta does not build foundation models and does not own distribution at the enterprise. It sits in the Enabler position: connecting to Microsoft Copilot, Amazon Bedrock AgentCore, Salesforce Agentforce, GitHub Copilot, Claude Code, and Codex. The company has placed itself at the measurement layer of that stack rather than the model layer, and Dell Technologies Capital’s Raman Khanna describes the target as “the system of record for AI value creation.” That is the FDE Enabler position in its clearest form.
What Is Concrete
Three products are named. Atlas measures AI value, adoption, and return. Forge covers how engineering teams use coding agents. Convoy serves organizations that provision their own AI capacity.
The platform connects to Microsoft’s Copilot suite, Amazon Bedrock AgentCore, and Salesforce Agentforce. On the coding-agent side, integrations include GitHub Copilot, Claude Code, and Codex.
Named customers are Atos and Wipro, alongside unnamed global insurance carriers. Florin Rotar, group chief technology officer and chief AI officer at Atos, is quoted on moving agentic AI from pilot to production. Partners listed include Microsoft, AWS, IBM, Slalom, and Trace3.
What is not established and therefore absent here: valuation, revenue, customer count, pricing, and how any of the three claimed percentages below was calculated.
The Numbers — Printed as Claims
The release contains three performance figures. They are printed here because the company published them. They are printed as claims because that is what they are.
Ascerta says its platform has improved ROI on AI initiatives by 47 percent, reduced agent launch times by 24 percent, and cut wasted AI spend by 86 percent across its customers.
Nothing is attached to any of them. No sample size. No time period. No named customer. No methodology for how a percentage improvement in ROI was computed.
An 86 percent reduction in wasted spend is a large assertion. The release gives a reader no way to test it. That matters more than usual here, because the product being sold is measurement. A measurement vendor publishing unmeasurable claims is worth noting without drawing a conclusion from it.
Three Implications
The Measurement Layer Is Now a Category
Sixteen months ago, AI costmanagement was a plausible enough category to raise a $4.9 million seed. Today, Dell Technologies Capital and Hitachi Ventures are betting $18 million that the real category is AI value measurement. When tier-one corporate VCs move in, the category definition tends to solidify fast — and the company that names the category often owns the definition for years.
The Pilot-to-Production Gap Is Where Budgets Stall
Florin Rotar at Atos being quoted on moving agentic AI from pilot to production is not incidental. That is the specific friction Ascerta is selling against. Enterprises have hundreds of AI pilots. The decision about which ones to scale is a budget decision. Budget decisions require evidence. Whoever provides that evidence gets embedded in the process — and embedded vendors are hard to displace.
Attribution Is the Product, and It Has Never Been Easy to Sell
The core technical claim is that Ascerta can tie an AI use case to the KPI it was meant to move. Attribution products are among the hardest to sell in enterprisesoftware — because the moment attribution shows a project is not working, someone’s budget is at risk. The buyers who most need honest measurement are often the same people least inclined to adopt it. That is the market dynamic sitting underneath the positioning, and it is not new.
Business Engineer Framework
The FDE Framework — Founders, Distributors, Enablers
Ascerta sits squarely in the Enabler position of the AI stack: it does not build models and does not own distribution, but it routes value signals between the two. The FDE Framework maps which layer of the AI economy a company actually occupies — and why Enablers can build durable margins even when model costs fall to near zero. The Map of AI covers where measurement platforms fit across all nine layers of the stack.
Pay-i renamed itself Ascerta because the question enterprises are willing to pay to answer changed. Cost was last year’s problem. Value — specifically, which AI bets are worth doubling down on — is this year’s. The rebrand, the $18 million, and the Dell Technologies Capital lead are all consistent with that reading. What is not yet consistent is a measurement vendor that publishes three large performance claims with no sample size, no time period, and no methodology attached. The company’s entire value proposition is making AI outcomes legible. That standard applies to its own numbers too.
Source: GlobeNewswire — Ascerta Series A announcement, September 30, 2026. Wire release read as mirrored full text; globenewswire.com does not serve this publication and ascerta.ai carries no funding post at time of writing. Nothing in this article is investment advice.
91,000+ executives read Business Engineer for the AI strategy frameworks cited by ChatGPT, Claude, and Perplexity.
Everything above is taken from Ascerta’s own Series A announcement of 30 September 2026, distributed through GlobeNewswire. That site does not serve this publication and ascerta.ai carries no funding post, so the release was read as a mirrored full text. Nothing has been independently verified.The reported improvements — 47 per cent better return on AI initiatives, 24 per cent faster agent launches and 86 per cent less wasted AI spend — are the company’s own figures for its own product.
The announcement gives no sample size, no time period, no named customer and no methodology for any of them. They appear above as claims and should be read as claims.Nothing above presents those figures as demonstrated results, or as evidence that the product works as described. The statement that the platform ties use cases to the business measures they were meant to move is likewise the company’s description of its own software.
The reading of the rename from Pay-i to Ascerta as a shift from cost measurement to value measurement is this publication’s analysis of one company’s repositioning. It is not a survey of the category and should not be read as one.Not established and therefore absent: any valuation, revenue, customer count or pricing. Nothing above predicts anything, and nothing here is investment advice.
Gennaro is the creator of FourWeekMBA, which reached about four million business people, comprising C-level executives, investors, analysts, product managers, and aspiring digital entrepreneurs in 2022 alone | He is also Director of Sales for a high-tech scaleup in the AI Industry | In 2012, Gennaro earned an International MBA with emphasis on Corporate Finance and Business Strategy.
Scroll to Top
Discover more from FourWeekMBA
Subscribe now to keep reading and get access to the full archive.
