Nebuly Raises $21M Series A to Measure Enterprise AI ROI

Iveco Group had dozens of internal AI copilots to understand. In Nebuly's account of the deployment, surveys and interviews left the automotive manufacturer's teams with a fragmented view of how those tools were being used. The software was spreading through the business faster than the explanations could travel back to the people responsible for it.

Nebuly, headquartered in Turin and New York, is expanding that measurement business with a $21M Series A announced November 12, 2026. Planven led the round, which will support US and European go-to-market expansion and further platform development. The company connects the work employees and customers do through AI agents with estimates and signals that managers can use to assess its value.

The commercial opportunity sits in the conversation between the team deploying AI and the people deciding what the business can afford. Nebuly makes the assumptions behind its calculations visible and adjustable, according to its announcement. That feature puts a human disagreement inside the product: finance and operating teams can examine how an estimate was assembled before it becomes part of a budget or workforce decision.

What Nebuly's Series A will fund

Nebuly's round includes Exor Ventures, Tenacity Capital, TGVG Capital and Fastweb+Vodafone through Founders Factory, alongside returning investors and angels. CDP Venture Capital SGR also supported the financing through its Artificial Intelligence Fund and Digital Transition Fund, the latter identified with Italy's PNRR and the European Union's NextGenerationEU program. The announcement does not provide a valuation or enough prior-round detail to establish total funding to date.

The founding team spans the technical and commercial sides of the product. Francesco Signorato is CEO, Diego Fiori is CTO and Julien Roux is CRO. Founders Factory's investment account describes Signorato's AI work at MIT and IBM, Fiori's experience on Giotto.ai's founding team, and Roux's earlier investment work at Exor. Their current assignment connects models, enterprise deployments and the people who must find a business reason to keep paying for them.

Nebuly says the new capital will grow its commercial organization across the US and Europe while funding platform development. Expansion carries a particular sales requirement for a measurement business. A prospective customer needs confidence in the calculations as well as the software producing them, especially when several departments will use the results to argue for different priorities. Explaining the method becomes part of earning the contract.

How Nebuly measures enterprise AI work

Nebuly's documentation divides the platform into Experience, ROI and Proficiency. Experience examines employee and customer interactions for intent, feedback, friction and risk signals. ROI and Proficiency currently focus on employee agents, addressing estimated economic value, spending, workforce capability and opportunities for training. Those product boundaries help explain what kind of evidence a buyer should expect from each view.

An employee assistant might help complete a task faster, making time available for other work. Converting that effect into a financial estimate requires assumptions about how long the task would otherwise take and the value assigned to that time. The published product description brings estimated time value together with AI costs. A manager then has to decide how to use the available capacity, whether that means handling more customer work, improving quality or changing a team's plans.

Customer-facing agents create a different measurement problem. A conversation can reveal dissatisfaction, a cancellation risk or interest in an additional product. Nebuly describes tracking these commercial signals and revenue influenced by AI. Attributing an eventual sale requires attention to the other things that happened around it, including human service and existing customer intent. Treating those signals carefully makes the output more useful to the people deciding what to change in the customer experience.

Why the calculation needs an owner

Francesco Signorato links Nebuly's mission to decisions about productivity and headcount in the company's announcement. The connection gives AI measurement consequences beyond the technology budget. An executive who accepts a time-saving estimate may use it to reconsider hiring, service capacity or training, bringing the estimate into decisions that employees and customers will feel directly.

Visible assumptions give those executives something specific to inspect. Finance can question the value assigned to time, while an operating team can explain what successful completion means for a particular task. This is an interpretation of the product's commercial role, rather than evidence that every customer already works this way. It identifies the conversation Nebuly is trying to support: several functions agreeing on what the same AI activity is worth.

The platform also examines how effectively employees use AI and where upskilling could help. That connects measurement to an intervention the organization can actually make. A team encountering poor results may need better context, a different workflow or more training; a usage count alone leaves those choices difficult to distinguish. The value of the resulting evidence depends on whether managers can connect it to a change in the working day.

The enterprise relationship behind the dashboard

Nebuly reports multiple six-figure contracts and names Oura, Iveco Group, Fastweb+Vodafone, CNH and Zeiss among its customers. Those are company-reported commercial claims. The Iveco case study provides a more concrete picture of the deployment problem: understanding interactions across dozens of copilots and sharing findings with stakeholders who were previously relying on limited feedback. Its feedback metrics should be read as analytics coverage, rather than an equivalent increase in productivity.

The company says it connects to custom-built tools and assistants based on Claude, ChatGPT and Microsoft Copilot, and supports on-premise or private-cloud deployment. That gives Nebuly a position across an enterprise's existing AI estate. A buyer can examine activity across tools while retaining a choice about infrastructure, although the deployment described for any individual customer still needs to be checked on its own terms.

DevCuration's reporting on AI agent memory at Graphon Labs and AI governance at WhiteMatter covers neighboring parts of the enterprise agent market. Nebuly approaches that market through the evidence managers use to assess the work, making measurement another part of the relationship between the enterprise and its AI suppliers.

The Series A finances a wider commercial effort around that relationship. Nebuly's team will be selling to enterprises where an AI program can involve technical leaders, finance, product managers and the people responsible for workforce development. As those customers decide which agents to expand and where employees need support, the calculation on Nebuly's screen will keep returning to the same practical question: who can use the capacity it describes, and for what work?