Company Spotlight: Ascerta Makes AI Value Visible

Ascerta is building an operating layer for a problem that becomes more expensive as enterprise AI succeeds: knowing which systems create business value, which merely consume budget, and which deserve to scale.

The Bellevue, Washington company began as Pay-i with a focus on AI cost management. Its new name reflects a broader mandate. Ascerta now connects model activity, infrastructure spending, employee adoption, and business KPIs so enterprise leaders can judge AI as an investment portfolio rather than a collection of experiments.

That distinction matters because lower model prices do not automatically produce lower AI bills. Agents can call several models and tools, retry work, loop through reasoning steps, and run continuously. A cheaper token can still sit inside a costly workflow. Ascerta is designed to show the complete operating picture and connect it to outcomes a finance or business leader can recognize.

From Cost Tracking to an AI System of Record

Traditional cloud cost tools are good at answering where infrastructure spend went. AI programs create a harder question: what did that spend accomplish?

Ascerta instruments activity across providers and organizes it around applications, use cases, teams, and business units. Its documentation describes each provider call as a request that can be enriched with usage, latency, cost, success data, and business context. Teams can then define custom KPIs and value policies that translate technical activity into measures such as hours saved, revenue generated, conversion lift, or customer satisfaction.

That creates a common record for engineering, finance, product, and executive teams. Engineers can inspect models, latency, and failures. Product owners can compare use-case performance. Finance can see actual spend and unit economics. Executives can decide where to expand or stop investment.

Atlas, Forge, and Convoy

Ascerta has organized the platform around three operating missions.

Atlas gives leaders a portfolio view of AI initiatives. It measures returns by use case or business unit, exposes wasted spend, and helps teams compare opportunities on business value instead of technical novelty.

Forge focuses on AI coding agents and the new software development lifecycle around them. It helps engineering leaders understand adoption, guide behaviors, and distinguish activity from productivity. That matters as coding assistants move from occasional suggestions to agents that plan, execute, and revise work across longer tasks.

Convoy addresses provisioned AI capacity. Reserved throughput can improve predictability, but it also creates the risk of paying for idle capacity or placing the wrong workloads on expensive commitments. Convoy is designed to show utilization in real time, identify consolidation opportunities, and determine which use cases can scale without disruption.

Together, the products make Ascerta broader than a model-cost dashboard. The architecture follows the decisions enterprises actually face: which initiatives create value, how employees use AI in critical workflows, and whether the underlying capacity is allocated well.

A Microsoft-Shaped Founding Team

Ascerta's leadership has deep experience with enterprise software at scale. Co-founder and CEO David Tepper spent more than two decades at Microsoft and later led work around the company's internal generative AI strategy. Co-founder and CTO Doron Holan spent 27 years at Microsoft, where Ascerta says he helped architect throttling systems that handled hundreds of billions of requests each day, including OpenAI traffic. Co-founder and COO Erik Winters brings experience in growth, go-to-market, and analytics at early-stage companies.

That mix maps closely to Ascerta's problem. Measuring AI value requires infrastructure that can capture high-volume technical events, a business model that can translate them into decisions, and a product that multiple enterprise functions will trust.

The founders also appear to understand that observability without economic context is incomplete. A model can be fast, accurate, and inexpensive per request while the use case around it still fails to create enough value. Ascerta's thesis begins where the technical dashboard stops.

Early Enterprise Evidence

Ascerta displays relationships across a broad enterprise ecosystem, including Microsoft, AWS, Atos, Wipro, Trace3, IBM, Slalom, Infosys, and KPMG. Its website cites company-reported outcomes including a 47% improvement in ROI for generative AI initiatives, a 24% reduction in time to launch agents, and an 86% reduction in wasted generative AI spend.

Those figures are not independent benchmarks, but they define the standard Ascerta wants buyers to use. The product is not only promising cleaner cost allocation. It is promising measurable changes in the economics and pace of AI programs.

The company's work with LangChain offers a concrete example. Engineering observability can show what an agent did, while Ascerta adds business KPIs and value policies that explain what the work was worth. That combination is especially relevant in financial services and other environments where a successful technical run may still have an unacceptable cost, risk, or operational result.

Capital for a Broader Mandate

On September 30, 2026, Ascerta announced an $18M Series A led by Dell Technologies Capital, with participation from Hitachi Ventures, BGV, Wipro Ventures, and earlier investors. The round brought total reported funding to $22.9M. The company did not disclose a valuation.

Ascerta says it plans to grow from about 15 people to roughly 40 by the end of 2027. That is a hiring and execution signal, not proof of scale by itself. No current official careers page or verified open-role list was available at the time of review, so the more important question is where the company applies that added capacity.

Product depth, integrations, enterprise delivery, and measurement rigor will all matter. Ascerta is asking leaders to trust one system as a record of AI economics. Earning that role requires accurate data and a defensible translation from technical events to business value.

What Ascerta Must Prove Next

The market is moving from AI access to AI accountability. Companies already have models, copilots, and agent pilots. What many still lack is a durable way to decide which investments work across changing providers, prices, teams, and use cases.

Ascerta has chosen a useful position in that transition. It is not trying to pick the winning model. It is building the measurement layer that helps enterprises keep making decisions as models and infrastructure change underneath them.

The next proof will come from repeatability. Ascerta must show that its KPI and value-policy framework can produce trusted comparisons across industries, that its product family remains coherent as it expands, and that reported customer outcomes hold up at larger scale.

If it succeeds, the company could become the place where an enterprise answers a deceptively simple question: not how much AI it used, but whether the AI was worth using.