Public procurement is not exactly the part of technology that usually goes viral.
Then again, most public-procurement startups don’t raise €6 million only months after launching, reach around 150 customers, claim €1 million in annual recurring revenue within six months, and persuade investors that bidding for government contracts could become a major vertical-AI category.
That’s what is happening with Cato, a Milan startup using AI to help companies find public tenders, decide which ones are worth pursuing and prepare the mountain of paperwork required to bid.
Its September 7, 2026 seed round brings total disclosed funding to €7.6 million.
And Cato is attacking a surprisingly enormous market: Italy alone recorded €309.7 billion worth of public contracts in 2025.
The question is whether Cato has actually built defensible software for that market—or whether investors are getting overly excited about yet another AI wrapper.
What’s Going On With Cato?
The immediate reason Cato is suddenly appearing in startup news is money.
On September 7, Cato announced a €6 million seed round led by Keen Venture Partners. Returning investors included Italian Founders Fund, Vento and Heartfelt, with other previous backers including Moonstone, BHeroes, Alecla7 and Nova Venture also reported around the financing. Strategic angels included founders associated with Lexroom, Sibill and Pillar.
Read Keen Venture Partners’ Cato investment announcement
Read Forbes Italia’s September 7 Cato funding report
The round followed a €1.6 million pre-seed in April 2026, led by Italian Founders Fund.
That financing by itself would not be particularly remarkable.
The customer trajectory is.
Cato publicly launched around mid-April with roughly 30 customers. By June, reports said it had passed 100. The September funding announcement put it at 150 active customers across sectors including medical, IT, services and construction. Vento says 10,000 tenders have already been managed through the platform.
Founder Andrea Zorzetto has also said Cato crossed €1 million ARR in six months. Treat that as a founder-reported figure rather than audited financial disclosure, but for an enterprise SaaS company this young, it is still worth paying attention to.
That is the real story.
Cato appears to have found a workflow companies already spend considerable time and money performing—and one where customers may have a measurable reason to pay for automation.
What Is Cato?
Cato is an AI platform for companies that sell goods and services to the Italian public sector.
Its customers are not government agencies buying things.
They are the suppliers competing to win those government contracts.
A company might sell hospital equipment, IT services, construction work, cleaning contracts, office products or dozens of other things purchased through public tenders.
Traditionally, a bid team has to:
find the tender, download documents, read the specifications, check whether the company meets every requirement, identify deadlines, prepare certifications and declarations, produce a technical proposal, calculate the commercial offer and track everything across email, documents and spreadsheets.
Cato attempts to place much of that workflow inside one platform.
Its website says it can monitor procurement sources, prioritize opportunities against a company’s profile, read tender documents, identify requirements and inconsistencies, fill administrative paperwork, help draft technical and financial documents, track team responsibilities and analyze previous tenders and competitors.
Visit Cato’s official platform website
That makes it closer to a CRM + research system + document workspace + AI bid assistant than a simple chatbot.
Who Founded Cato?
Cato was co-founded by Andrea Zorzetto and Matteo Bossolini.
The pair met in June 2025 and began experimenting with the concept shortly afterward. Cato was formed later in 2025 and publicly launched in 2026.
Zorzetto brings the commercial and public-policy side.
He has said he previously worked with the UK Treasury and the City of Paris, and previously founded PeopleRank, a startup that ultimately failed. He had also worked around Plug and Play’s Italian operation. What makes his background relevant is that procurement is partly a software problem and partly a government-process problem.
Bossolini is the technical co-founder and CTO. Forbes reports that he began programming at 12, launched his first startup at 17 and had product experience across several startups before Cato.
Their pairing makes sense for the market: one founder obsessed with selling and understanding institutions, one founder shipping the AI product.
Why Is Cato Suddenly Getting So Much Attention?
Several signals have arrived almost simultaneously.
First is the funding velocity: €1.6 million in April followed by another €6 million approximately five months later.
Second is the customer growth: around 30 customers at launch to roughly 150 by summer.
Third is the reported revenue.
Zorzetto says Cato reached €1 million ARR in six months, faster than the company itself expected.
Fourth is usage.
Investor Vento says the company has now handled 10,000 tenders. Cato’s own site, which appears to display an older set of counters, shows more than 100,000 tenders filtered, 2,000 analyzed and 500 forms automatically completed. Those numbers measure different stages of the workflow, so they are not necessarily contradictory.
Fifth is the market itself.
ANAC says Italian public contracts reached €309.7 billion in 2025, up 13.9% in value from 2024.
See ANAC’s official 2025 Italian procurement figures
Cato is therefore not creating demand for public procurement.
It is trying to tax the inefficiency surrounding an existing gigantic market.
That is a much more compelling startup thesis.
AI Tribune readers who have followed companies such as Instinct, AfterQuery and Naïve will recognize the broader pattern: instead of building another general chatbot, startups are increasingly trying to own one valuable workflow end to end.
Cato has picked an unusually bureaucratic one.
How Does Cato Work?
The basic workflow looks like this:
Company data → tender discovery → AI analysis → eligibility/risk checking → bid drafting → human review → submission → historical intelligence
The first layer is procurement data.
Cato says it monitors more than 25,000–27,000 sources, including Italian procurement portals and contracting authorities.
It then matches opportunities against information about the customer’s company: sector, certifications, prior work, product catalogues and other relevant criteria.
When a tender looks promising, Cato reads the accompanying documentation and extracts things such as deadlines, eligibility rules, CPV codes, evaluation criteria and inconsistencies that may require clarification.
The agent can then help prepare administrative paperwork and draft technical or commercial material using information the company has already uploaded.
The final decision remains with the customer.
Cato explicitly positions the system as a copilot, not an autonomous bidder that quietly submits government contracts on your behalf.
What models does Cato use?
This is where the story becomes more interesting.
Cato does not appear to be building its own foundation model.
Its April 2026 privacy policy identifies OpenAI, Anthropic, Google and Mistral as AI providers used for intelligent analysis, data extraction and content generation. It also lists LlamaIndex for document parsing/indexing, AWS for cloud infrastructure, Supabase and Elasticsearch.
Read Cato’s privacy policy and technical supplier disclosures
That means Cato’s potential moat is probably not “we have a better LLM than OpenAI.”
It is the system around the models:
the procurement data, legal structure, company context, workflow, historical bids, competitor information, tender-specific interfaces and accumulated customer knowledge.
That is arguably the more defensible place for a vertical-AI startup anyway.
What Can You Actually Do With It?
Imagine a medical-device company selling surgical equipment.
Instead of an employee checking thousands of tender notices manually, Cato can identify tenders that appear relevant to the company’s actual catalogue and qualifications.
Then someone can ask the system to:
- explain the eligibility requirements;
- identify which lots match the company’s products;
- flag conflicting deadlines or clauses;
- identify required certifications;
- compare technical requirements against product specifications;
- pre-fill recurring administrative declarations;
- draft the structure of a technical proposal;
- show how competing suppliers have performed in previous tenders;
- retrieve previous company material that can be reused.
Cato claims one customer went from handling roughly one tender per week to three or four with the same team, while another reportedly identified €14 million in tender lots it was close to missing. Those are company-published customer testimonials, not independently verified case studies.
That distinction matters.
But the use case itself is very concrete.
Cato Pricing
Cato currently does not publish standard commercial pricing on its website.
| Option | Current status |
|---|---|
| Free plan | No public free commercial tier found |
| Free trial | No general trial publicly advertised |
| Paid plan | Pricing not publicly listed |
| Enterprise | Demo/contact sales |
| API pricing | Not publicly listed |
| Demo | Available |
| Public tender search | Some tender-discovery functionality is publicly surfaced |
The company’s main call to action is to book a demo.
That opacity is understandable for early B2B software with very different customer sizes, but it also makes independent comparisons harder.
Competitors such as TenderEU and Tendify already publish pricing publicly, which gives buyers a useful benchmark.
The Numbers Behind the Hype
| Metric | Reported figure |
|---|---|
| Founded | Late 2025 |
| HQ | Milan, Italy |
| Public launch | April 2026 |
| Pre-seed | €1.6M |
| Latest seed | €6M |
| Total disclosed funding | €7.6M |
| Valuation | Not disclosed |
| Active customers | ~150 |
| Tenders managed | 10,000+ reported by investor Vento |
| ARR | €1M claimed by founder |
| Team | ~35 people |
| Italian procurement market, 2025 | €309.7B |
Sources: Keen Venture Partners, Forbes Italia, Vento, Andrea Zorzetto and ANAC.
What Makes Cato Different?
Cato’s strongest idea is not “AI writes proposals.”
Plenty of products can generate text.
The more interesting part is context accumulation.
A tender platform becomes more useful if it remembers what the company sells, what certifications it has, which tenders it previously entered, how those bids scored, who competed, what documents were used and what requirements repeatedly appear.
That turns Cato into something closer to a system of record for public-sector revenue.
Keen Venture Partners explicitly describes the long-term opportunity in similar terms: a kind of CRM for public-sector sales.
The other advantage is geography.
Cato began in Italy, which is messy.
That sounds like a disadvantage until you consider that solving fragmented portals, local regulatory requirements and complex documentation may create more product depth than starting with an easier procurement environment.
Cato also integrated the technology of Avvista.ai in June through an acquihire, adding a self-service approach aimed at companies without dedicated tender departments.
If it can eventually generalize what it learns in Italy across Europe, the difficult home market could become useful training.
Who Is Cato Competing Against?
Cato does not have the category to itself.
| Platform | Main angle | Main difference vs Cato |
|---|---|---|
| Cato | Italian public-tender workflow | Deep Italy-first procurement workflow |
| Stotles | Public-sector sales + bid intelligence | Stronger emphasis on broader B2G pipeline and pre-tender intelligence |
| Tendersight | European tender lifecycle | Broader 27-country positioning |
| TenderEU | EU-wide discovery, scoring and filling | 27 EU states, 24 languages, public pricing |
| Tendify | Tender discovery + AI briefs | Simpler monitoring-focused product and transparent pricing |
Stotles now positions itself as an “AI operating system” for government sales, with tender discovery, buyer intelligence, bid qualification and drafting.
See Stotles’ AI public-sector platform
Tendersight says it covers public tenders across 27 European countries and combines discovery, drafting, compliance and performance tracking.
TenderEU is particularly interesting because it advertises all 27 EU member states, 24 languages, EU-hosting and public prices from €200 per month for basic search to €3,000 per month for bid filling.
See TenderEU’s platform and pricing
And Tendify advertises monitoring across European markets with a currently listed annual plan.
So Cato’s challenge is not proving AI can help with procurement.
It is proving that its Italian workflow depth, customer data and execution speed can become more defensible than competitors’ broader geographic coverage.
What People Like About Cato
Independent review data is still thin.
There is no mature Reddit, G2 or Hacker News consensus around Cato yet, which is unsurprising given its age and B2B niche.
The positive evidence mostly comes from customers quoted by Cato and its investors.
Recurring themes are straightforward:
time savings, fewer documents read manually, the ability to manage more tenders with the same team and having company-specific context available instead of starting from a blank ChatGPT session every time.
Cato’s website displays companies including SOL Group, CNS, MOVI, AHSI, Favero Health Projects and othersamong organizations using the platform.
The strongest independent-ish signal is not a testimonial, though.
It’s retention-style behavior.
Keen says customers open Cato every working day, while an earlier founder interview described customers complaining quickly when the agent was temporarily unavailable because it had become part of their tender workflow. Investor claims obviously have incentives attached, but daily-use software tends to be more meaningful than a flashy demo.
What People Don’t Like—or Should Be Skeptical About
Cato is new enough that there is not yet a substantial public body of negative customer reviews.
That does not mean there are no drawbacks.
Pricing is opaque
There is no published pricing table, making it difficult for small companies to know whether Cato is economical before entering a sales process.
Direct integrations appear limited
The company’s April privacy policy says Cato did not then support direct API integrations with third-party applications. Data transfer relied on file imports/exports and automated collection of public procurement data.
That could change quickly, but it is a notable limitation for companies with mature CRM, ERP or document systems.
It depends on third-party AI models
OpenAI, Anthropic, Google and Mistral appear in Cato’s supplier list.
If Cato’s value becomes mostly prompt orchestration, competitors can reproduce that.
Its moat therefore depends on the proprietary workflow, data layer and customer history becoming substantially more valuable than the underlying models.
Some marketing language deserves skepticism
Cato’s website claims its consistency checks can “eliminate human error.”
In a high-stakes legal and procurement workflow, that is too absolute.
AI can reduce certain errors. It should not be assumed to eliminate them.
Europe is harder than Italy
Italy may be an excellent wedge.
But procurement law, portals, languages, commercial practices and document structures vary significantly across Europe. Expanding the product could be more difficult than simply turning on multilingual models.
Privacy and Security
This is one of the more important sections of the review because companies may upload sensitive commercial documents, pricing, certifications and financial information into Cato.
There are some encouraging signals.
Cato obtained ISO/IEC 27001:2022 certification, with DNV’s certificate dated August 28, 2026 and valid through August 2029.
View Cato’s ISO 27001 certificate
Cato also says customer and tender data are not used to train general-purpose AI models, and its privacy policy states customer environments are logically segregated.
The privacy policy lists encrypted HTTPS/TLS communications, daily backups, access monitoring and permission-based access controls.
There is, however, an important nuance.
Cato’s marketing site says data stays within the EU. Its detailed privacy policy says providers such as OpenAI, Anthropic and Google may process data outside the European Economic Area, using contractual safeguards, while an EU-focused option using Mistral is available for greater localization.
That distinction should matter to enterprise buyers.
Retention is also documented: company data may remain for the contract period plus 12 months for stated legal/fiscal purposes, account deletion is targeted within 30 days, and backups can persist for up to 90 days following deletion.
So the security posture looks unusually documented for a startup of this age—but buyers with sensitive bids should still review the contractual setup and model-provider configuration rather than assuming “private AI” means no external model processor ever touches their data.
Is Cato Actually Legit?
Yes, in the basic meaning of the word.
Cato is a real incorporated company operating as AZMB Srl in Milan. It has identifiable founders, institutional venture backing, a functioning commercial platform, publicly referenced customers, documented funding and ISO certification.
Its latest €6 million financing involved established European investors and legal advisers.
But “legit” does not mean Cato is automatically worth buying.
It also does not validate the company’s internal revenue numbers, guarantee AI accuracy, prove that every customer gets the advertised productivity gains or tell us whether Cato can defend itself against larger European competitors.
Those are different questions.
Cato AI Pros and Cons
| Pros | Cons |
|---|---|
| Solves an expensive, highly repetitive workflow | Public pricing is unavailable |
| Fast reported customer growth | Still extremely young |
| €7.6M in disclosed funding | No disclosed valuation |
| €1M ARR claimed within six months | ARR is founder-reported |
| Deep specialization in Italian procurement | Geographic specialization may complicate European expansion |
| Uses company history and tender data, not just generic prompts | Relies on third-party foundation-model providers |
| ISO 27001 certified | Some AI processing may occur outside EEA depending on provider |
| Human keeps control of final submission | Direct API integrations were not supported in the April privacy policy |
| Potential system-of-record moat | Limited independent user-review data |
So… Does Cato Deserve the Hype?
More than most AI startups this young.
The reason is not that Cato has invented a new model.
It hasn’t.
The reason is that it appears to have found a high-value workflow where current AI models finally make automation practical.
Tender documentation has almost everything LLM-era software likes: enormous amounts of text, repetitive formats, structured requirements, historical documents, deadlines, searchable public data and labor-intensive drafting.
And the person using the software has an extremely obvious ROI calculation.
If Cato lets a five-person bid team pursue three times as many legitimate opportunities without tripling headcount, it does not need to become magical AGI.
It just needs to work reliably.
The early traction is therefore much more interesting than the €6 million headline.
Still, there are unanswered questions.
Can the company maintain its reported growth after the easiest early adopters are signed?
Will €1 million ARR become €5 million or €10 million without exploding sales costs?
Can Cato expand outside Italy?
Will customers trust AI-generated material enough for high-value contracts?
And can its accumulated procurement intelligence become a genuine moat before Stotles, Tendersight, TenderEU or another competitor closes the gap?
For now, the underlying company looks more interesting than the hype around it.
That is usually a good sign.
Who Should Try Cato?
Cato looks most relevant for:
- Italian SMEs regularly competing for government work
- Medical-device and healthcare suppliers with document-heavy tenders
- Construction companies pursuing public works
- IT and professional-services firms with dedicated bid teams
- Larger companies still coordinating tenders through email, Excel and shared folders
Companies that rarely participate in public tenders should probably skip it unless Cato’s eventual self-service pricing makes occasional use economical.
Likewise, companies operating almost entirely outside Italy should compare broader-European platforms first.
FAQ
What is Cato AI?
Cato is a Milan-based AI platform that helps companies discover, analyze, prepare and manage bids for public-sector tenders.
Who founded Cato?
Cato was founded by Andrea Zorzetto and Matteo Bossolini.
How much funding has Cato raised?
Cato has disclosed €7.6 million, comprising a €1.6 million pre-seed and a €6 million seed round led by Keen Venture Partners.
What is Cato worth?
No official valuation was disclosed with its September 2026 seed round.
Is Cato AI free?
Cato does not currently publish a general free commercial plan or standard price list. Customers are directed to request a demo.
Does Cato use OpenAI?
Cato’s April 2026 privacy policy lists OpenAI, Anthropic, Google and Mistral as providers used for AI analysis, extraction and content generation.
What are the main Cato alternatives?
Relevant competitors include Stotles, Tendersight, TenderEU and Tendify, depending on geography and whether a company needs tender discovery, broader public-sector sales intelligence or full bid preparation.
Final Take
Cato is a good example of where the next wave of serious AI companies may come from.
Not another chatbot.
Not another image generator.
A piece of obscure, expensive, bureaucratic business infrastructure nobody outside the industry thinks about—until AI suddenly makes much of it automatable.
The next milestone to watch is not another funding round.
Watch whether €1 million ARR becomes several million, whether the customer base keeps accelerating past 150, and especially whether Cato starts expanding into other European procurement systems without sacrificing the domain depth that currently makes it interesting.
If that works, Cato could become considerably more important than its current €6 million seed round suggests.
If it cannot, it may remain an impressive Italian vertical SaaS company in an increasingly competitive category.
Would you trust AI to help prepare a multimillion-euro government bid—or is that one workflow where humans should stay firmly in control?
Leave a Reply