Nebuly Raises $21 Million to Measure Enterprise AI’s Business Value
- Cargo
- Machine Learning
- Shipped
- Shipper
- Mike Wheatley
- Readers
- 578
Nebuly’s platform tracks how AI agents affect employee capacity, revenue and customer churn, giving large companies a way to assess returns.

Nebuly has raised $21 million to help large companies measure what their artificial intelligence systems contribute to business results. Its platform tracks the work performed by AI agents and connects that activity to measures such as employee capacity, revenue and customer churn. The aim is to give business and finance teams evidence they can use to assess which projects are delivering value.
The funding round was led by Planven, with participation from Exor Ventures, Tenacity Capital, TGVG Capital, Fastweb, Vodafone and CDP Venture Capital’s Digital Transition Fund. Nebuly plans to use the financing to expand its go-to-market hiring in the United States. Most of its customers are currently based in Europe.
Usage counts don’t show business impact
Companies have adopted a mix of commercial AI tools and internally developed applications, including products from OpenAI, Anthropic, Microsoft and Google. That patchwork makes it difficult to compare projects and determine which ones produce measurable results. Counts of prompts, responses or tokens can show that employees are using AI, but they don’t establish whether the work saves time, improves service or generates revenue.
Nebuly co-founder and Chief Executive Francesco Signorato said companies need better answers as they face questions about the value of their AI spending. He described usage-focused measurement as “tokenmaxxing,” where activity itself becomes the goal. A high prompt count, he said, doesn’t show whether an employee got an hour back or a customer received the answer they needed.
The distinction matters as companies expand their spending. A Gartner survey cited in the source found that 85% of business leaders planned to increase AI spending in the following year, while 22% said they had scaled the technology across multiple business units. The figures point to a gap between investment plans and broad deployment. The source also reported that Uber had used its entire budget for AI in 2026 within four months, an example of how quickly spending can run ahead of planning.
Signorato’s concern extends beyond project accounting. If companies make staffing and head-count decisions based on assumed productivity gains, weak measurement can feed those assumptions without showing whether AI has actually changed the amount or quality of work people can do. In his view, decisions of that scale call for evidence about outcomes.
Measuring internal and customer-facing agents
Nebuly’s platform connects to AI agents in a customer’s information technology environment and analyzes their behavior through a unified control plane. For employee-facing agents, it measures operational effects such as hours saved and capacity freed for other tasks. It also tracks work that teams can take on with AI that they couldn’t previously do.
The platform categorizes the kinds of tasks employees assign to agents, from routine work to higher-level strategic tasks. That view can help companies see where their tools are being applied and where they appear most useful. It also gives teams a basis for comparing the agent’s role across different kinds of work, beyond simply recording how often it is used.
For customer-facing agents, Nebuly tracks their relationship to revenue, customer churn and upsell opportunities. These measures are intended to connect agent activity with commercial outcomes. The source doesn’t report benchmark results or quantified customer gains, so the platform’s described capabilities are measurement categories rather than evidence of a particular return.
Financial review and deployment choices
Because financial conclusions depend on assumptions, Nebuly says its platform lets finance teams inspect the calculations behind its performance insights. Teams can review and adjust the financial assumptions used in the analysis, making it possible to examine how the results were produced. That auditability is relevant when estimates of saved time or added capacity feed into decisions about budgets and staffing.
Customers can deploy the platform on premises or in private cloud environments to keep sensitive business data within their preferred infrastructure. Those options may matter to large organizations that need measurement across multiple systems while limiting where business data is processed.
Large enterprises are the initial market
Nebuly says almost all its customers generate more than $100 million in annual revenue, and around half exceed $1 billion. Named customers include Carl Zeiss, Iveco Group and Oura Health. The concentration among large organizations fits a product aimed at companies managing multiple AI tools, business units and financial priorities.
Planven’s Giovanni Canetta Roeder said the firm backed Nebuly because large organizations will need a way to measure returns from AI. He characterized the product as turning what agents do into evidence of business value, pointing to the company’s enterprise traction as a factor in the investment.
The test for Nebuly will be whether its measurements help customers distinguish visible AI activity from outcomes that matter to the business. Tracking hours, capacity, revenue and churn gives companies a more practical starting point than prompt totals; the value of the platform will depend on how clearly its calculations connect agent behavior to results that finance and operating teams can verify.