Nebuly raises $21M to prove what AI agents are worth to the board

Nebuly, a startup whose software measures the business value of enterprise AI agents, has raised a $21 million Series A led by Planven, according to Techmeme’s summary of a Tech.eu report. The platform analyzes how agents affect productivity and revenue.

What the platform measures, and where it runs

Tech.eu, which first covered the round on November 12, describes a platform that pulls usage data from two kinds of deployments: tools employees use internally and assistants that face customers. We could not find a press release from Nebuly or its investors, so the details below rest on that report.

For internal tools, the product tracks efficiency gains such as time saved and added capacity, plus new tasks staff can now perform. It also shows how teams use AI, whether the work is high value, and where training could help. For customer-facing assistants, it measures revenue that AI influenced and flags churn risks and upselling opportunities.

Finance teams can review and change the assumptions behind the calculations, which matters when a number goes to a board. According to Tech.eu, Nebuly works with custom-built tools and assistants built on Claude, ChatGPT and Microsoft Copilot. It can run on-premises or in a customer’s private cloud, so company data is not sent to external model APIs.

Besides Planven, Tech.eu lists Exor Ventures, Tenacity Capital, TGVG Capital and Fastweb+Vodafone, the last via Founders Factory. Returning and angel investors also took part, as did CDP Venture Capital SGR through its Digital Transition Fund, which is backed by the EU’s NextGenerationEU programme. The money will go to a larger go-to-market team in the United States and Europe and to further development of the platform.

Co-founder and CEO Francesco Signorato framed the pitch around boardroom pressure: “Every enterprise is about to face the same question from its board: what is AI actually worth to us?”

Open questions behind the return-on-AI pitch

Several things are not established. Tech.eu names no customers, gives no valuation, and does not say how many companies use the product. The claims about productivity and revenue impact come from the company and have not been independently checked. How the platform separates revenue caused by an assistant from revenue that would have happened anyway is not explained, though the adjustable assumptions suggest that part is left to each customer’s finance team.

Some background helps. Public listings such as Crunchbase and Caplight describe Nebuly as a user analytics company for large language model products, founded in 2022, with Exor Ventures among its earlier backers. Caplight lists a pre-seed round in March 2022 as its previous financing, though these aggregator entries are not official and sources differ on its headquarters. The Series A therefore reflects a shift from tracking how people talk to chatbots toward the finance question of what agents return.

The round also fits a pattern in our recent coverage of funding for the plumbing around agents. Last week we reported that Hatz AI raised $24M to sell AI to small firms through their IT providers. Techmeme also logged a $20M Series A for WhiteMatter, which builds oversight agents for other AI agents, citing Business Insider. Each is a bet that once agents are deployed, companies will want to watch, govern and justify them.

What would show the pitch is working

The first test is whether Nebuly publishes named enterprise customers and measured results, especially on the revenue side, where attribution is hardest. A second is how much finance teams actually adjust the default assumptions. The company has said it will expand sales in both the US and Europe, so hiring and customer announcements in those regions will show how fast demand is growing. Pricing and plan details were not reported, so buyers cannot yet compare cost against the savings the tool claims to find.