Nebuly
by Nebuly
AI ROI platform that measures what employee and customer AI agents are worth to the business
Nebuly is an AI ROI platform for enterprises that want to know what their AI agents return. It reads the conversations employees and customers have with internal copilots and customer-facing assistants, then estimates time saved, revenue influenced and team proficiency, so CIOs and finance teams can decide which agents to scale, fix or retire.
Nebuly is an analytics platform from a Turin and New York company founded in 2022 by Francesco Signorato (CEO), Julien Roux and Diego Fiori. The company started with user analytics for large language models and has since moved into measuring the business value of AI agents. It connects to existing employee and customer agents without changes to the agents themselves, including custom tools built on Claude, ChatGPT and Microsoft Copilot, and maps each conversation to the task behind it. Three modules sit on that data. AI Experience reports resolution rate, user frustration, churn risk, revenue risk, upsell signals and unmet demand. AI ROI prices time saved on completed tasks at the cost of the people who would otherwise have done the work, subtracts AI running costs, and breaks the result down by team, tool and use case; finance teams can change the underlying assumptions. AI Proficiency scores teams from novice to expert based on how they actually use AI and recommends training. Nebuly removes personal identifiers before analysis and runs its own models inside the deployment, so conversation data is not sent to outside AI providers. It is offered as SaaS on a Nebuly-managed Azure tenant or self-hosted on Kubernetes in AWS, Azure, GCP or on premises. Named customers are Oura, Iveco Group, Fastweb+Vodafone, CNH and Zeiss. On November 12, 2026 the company announced a $21M Series A led by Planven, which it says will fund go-to-market teams in the US and Europe.
A CIO or head of AI who has rolled out several copilots and assistants and now has to show the board or CFO what that spend returns, team by team.
A per-team, per-tool estimate of hours saved and revenue influenced by AI agents, built from real conversation data with assumptions finance can adjust.
At a Glance
- Category
- Data & Analytics
- Pricing
- Contact for pricing, Subscription
- Target Market
- CIOs, CFOs, Heads of AI, Customer Experience Leaders
- Deployment
- Cloud-first, Self-hosted, Multi-cloud
- Founded
- 2022
- Headquarters
- Turin, Italy
Key Features
- ✓AI ROI
Prices time saved on completed tasks at the cost of the people who would have done them, minus AI running costs, by team, tool and use case.
- ✓AI Experience
Maps every agent conversation to its underlying task and reports resolution rate, frustration, churn risk, revenue risk, upsell signals and unmet demand.
- ✓AI Proficiency
Scores each team from novice to expert from observed AI usage instead of surveys, and recommends training to close the gaps.
- ✓No-change agent integration
Connects to existing employee and customer agents, including custom builds on Claude, ChatGPT and Microsoft Copilot, without modifying them.
- ✓Self-hosted deployment
Runs on Kubernetes in your own AWS, Azure or GCP account or on premises, so conversation logs never leave your network.
- ✓PII removal and in-deployment models
Strips names, emails and other identifiers before analysis and runs Nebuly's own models inside the deployment rather than calling external AI providers.
- ✓Adjustable finance assumptions
Lets finance teams change the cost and value assumptions behind ROI figures, so the numbers can be challenged and re-run.
Capabilities
Use Cases
- •Copilot portfolio review
An automotive group like Iveco measures dozens of internal copilots from one dashboard and decides which ones to expand or retire.
- •Board-level AI value reporting
A CIO reports hours saved and capacity freed by AI per business unit, with assumptions the CFO has reviewed and adjusted.
- •Customer assistant revenue tracking
A telco tracks AI-influenced revenue, churn risk and upsell signals in its customer assistant conversations to prioritise fixes and follow-ups.
- •Targeted AI training
An HR or transformation team uses proficiency scores to find the departments that use AI least well and sends training there first.
Ideal For
Best For
- ✓Enterprises running dozens of internal copilots that need one place to compare their usage and value
- ✓Finance and AI leaders building a business case to renew, expand or cut AI licences
- ✓Customer-experience teams tracking churn risk, frustration and upsell signals in assistant conversations
- ✓Regulated companies that need conversation analytics to stay inside their own cloud or data centre
- ✓L&D and transformation teams deciding which departments need AI training first
Not Ideal For
- ✗Teams that want a causal proof of AI ROI, since the figures rest on adjustable assumptions about what work AI replaces and what it is worth
- ✗Small companies with one or two AI tools, where a vendor's built-in usage dashboard is likely enough
- ✗Engineering teams looking for LLM tracing, prompt debugging or evaluation tooling, which is a different product category
- ✗Buyers who need published list pricing before a sales call
Integrations
Deployment
Market Analysis
Pros
- ✓Named enterprise customers in manufacturing, telecom and health tech (Oura, Iveco Group, Fastweb+Vodafone, CNH, Zeiss)
- ✓Self-hosted deployment on AWS, Azure, GCP or on premises keeps conversation data in the customer's network
- ✓Trust centre lists SOC 2 Type 2, ISO/IEC 27001, ISO/IEC 42001 and GDPR
- ✓Finance can edit the assumptions behind ROI numbers instead of accepting a black-box score
Cons
- ✗ROI output depends on judgment calls about what work AI replaces and what it is worth; time saved does not automatically cut costs, and influenced revenue does not prove causation (Magica, Nov 2026)
- ✗No independently verified customer return figures have been published
- ✗No public pricing, and the infrastructure cost of a private deployment is unspecified
- ✗Overlaps with Ascerta and with the value-measurement modules governance vendors now ship, so differentiation will be tested
- ✗No G2 or Capterra reviews and almost no Hacker News discussion of the current product; the company's older open-source optimisation projects are marked unmaintained
Pricing
Nebuly Cloud (SaaS)
Contact for pricing
- ✓Managed on Nebuly's Azure tenant
- ✓Continuous updates
- ✓SSO and RBAC
- ✓Slack integration
Nebuly Self-Hosted
Contact for pricing
- ✓Kubernetes on AWS, Azure, GCP or on premises
- ✓Conversation data stays in your network
- ✓Configurable for data-residency requirements
Nebuly publishes no list prices; access starts with a demo or a public playground. The company says it holds multiple six-figure contracts but discloses no terms. A self-hosted deployment also carries your own Kubernetes infrastructure and integration costs, which are not quantified anywhere public.
Security & Compliance
Connect
Sources
This page was written from 7 sources, 6 on domains other than nebuly.com.
Stay Ahead of the Curve
Weekly enterprise AI insights for technology leaders. No spam, no vendor pitches—unsubscribe anytime.
SubscribeRelated Products
Edelbyte
One engine for search, analytics, vector, graph and geospatial queries, built for AI agents that query company data constantly
Revefi
AI FinOps and observability agents that cut LLM, Snowflake, Databricks and BigQuery spend
Corvic AI
Agentic data engine that turns PDFs, images, sensor logs and tables into cited answers and pipelines
Snorkel AI
Expert-built training data, RL environments and evaluation benchmarks for frontier labs and enterprises building specialized AI agents