Mintive says 50 customers send it 2.5 million agent calls a day as it raises $11 million

Mintive, a company that sells software to control what AI agents can reach inside a business, said on October 29 that it has raised $11 million in total, including a $7 million round led by Cowboy Ventures, and that more than 50 paying customers now send 2.5 million agent calls a day through its platform, according to its launch announcement. The company was previously called MintMCP.

What happened

In the announcement, Chief Executive Jiquan Ngiam said the latest $7 million round was led by Cowboy Ventures, with Operator Collective and Burst Capital joining and Coatue, Hustle Fund and Maven Ventures taking part. Angel investors include Frederic Kerrest, co-founder of Okta, Jeff Dean, co-founder of Discovery Loop and former Chief Scientist at Google, and Andrej Karpathy, co-founder of OpenAI, the post said.

The company said it has "more than 50 paying customers," among them Ro, Modern Treasury, Braze, Coursera, the University of Michigan, Apax and Stability AI, and that it has handled "more than 200 million" agent calls in seven months. At the current rate of 2.5 million calls a day, a 30-day month would hold 75 million calls, which is 2.6 times the seven-month average of 28.6 million a month (computed from the company's figures).

How it started

The company describes its origin in one sentence: "We started Mintive after seeing the first enterprise agents enter production with broad access and no clear record of which agent had taken which actions." Its about page puts the problem this way: "Most people carry far more permissions than any one task needs, and an agent that borrows a person's token inherits all of them."

The founders are Ngiam and Vijay Vasudevan, according to TechCrunch. TechCrunch reported that Ngiam was the founding engineer of Coursera and later worked on neural networks at Google Brain, and that Vasudevan, the chief technology officer, is a co-creator of TensorFlow. The company's own launch post describes the team as "scientists and builders from Stanford, Google Brain, Coursera, Applied Intuition, Snorkel and Amazon." Neither the launch post nor the about page gives the year the company was founded; TechCrunch reported that it launched its product in February.

The product is built around one idea, which the company states as: "we treat each agent the way an enterprise treats an employee: it gets its own identity, only the access its work requires, and a record of everything it does, regardless of the model or tool underneath."

How it grew

The launch post lists five products: an MCP gateway that governs which tools agents can use, an LLM gateway that routes calls through approved models and sets budgets, agent identities with scoped and revocable credentials, an agent monitor that records the chain from person to agent to action, and coworker agents that work in Slack and Teams. MCP, the Model Context Protocol, is a standard that lets AI tools connect to company systems.

Three customer stories on the company's site show how those products were used. They are the company's own write-ups with customer quotes, not independent reporting.

At Coursera, the online education company, the story says AI adoption reached 100% but "no one could say exactly what was connected or where." About 2,000 employees are now on the system, with 70 or more MCP connections, and more than a quarter of AI usage comes from non-technical teams, the story says. Coursera's chief technology officer, Mustafa Furniturewala, is quoted as saying: "My concern was less about the technology itself and more about the lack of visibility and consistency across the company."

At Modern Treasury, the payments company, the story says deployment took less than three weeks and the team governs more than 70 MCP servers. Engineering leader Ankit Kumar is quoted as saying that a bank-operations agent "saves our Technical Account Managers 30 minutes to 4 hours per case." Kumar said he found Mintive after "knee-deep in POCs with a couple of other gateway companies."

At Stability AI, the story says the company moved "from evaluation to company-wide in less than a month," with more than 140 employees and more than 40 connectors. Head of Security Christian Burrows is quoted as saying: "The problem isn't getting people to use AI. It's that every workflow was held together by personal API tokens sitting in a dotenv file on someone's laptop."

What the founders and their customers say

The launch post, written by Ngiam, frames the market with a figure from the research firm Gartner: it "expects the average Global Fortune 500 company to run more than 150,000 AI agents by 2028, up from fewer than 15 in 2025," according to the post. Aileen Lee, founder and managing partner of Cowboy Ventures, is quoted in it as saying that the "major gating factors" for companies adopting AI are "governance, IT and security team ability to oversee, identity management, and security."

Burrows said in the Stability AI story: "A company doesn't exist because I'm trying to secure it. I'm trying to support the business." The company's pitch, in the words of its post, is that Mintive is "model-agnostic," so that "the identities, permissions and audit trail stay the same underneath" whichever model is used.

What's next

The company said it is hiring across engineering, marketing and go-to-market. Its stated direction is on its about page: "Further up, the way companies organize their work and software will change. We are building toward that from trusted access up."

TechCrunch described the field as crowded, saying agent management is offered by Amazon, Microsoft, OpenAI and Anthropic, as well as by many smaller companies, and that more than 40 startups from Y Combinator alone work on the idea. The company's published figures do not say how many of its 50 customers pay for more than one product or how large its revenue is.

Our view

Our view is that Mintive's numbers are early but specific. More than 50 paying customers and 2.5 million calls a day are the company's own figures, and the three customer stories are written by the company, so they show what Mintive chooses to publish rather than the full range of its customers. What they do show is a consistent description of the problem: companies adopted agents faster than they could track what those agents could reach.

The threat is the one TechCrunch names: large platform companies sell agent management as part of products customers already buy, and a company whose product sits between customers and those platforms depends on staying neutral. A second threat is that the revenue behind $11 million of funding is not disclosed. The opportunity is that the demand it describes comes from security and IT teams that, in its customers' words, could not say which tools were connected.

We would watch two milestones as conditions, not predictions: whether the company publishes revenue or customer counts beyond 50 in its next announcement, and whether the daily call volume it reports keeps rising faster than its customer count. If the call rate grows while customers stay near 50, the growth would be coming from deeper use by existing customers, which would be a different story from one of new customers. Funding is a milestone, not proof, and this one does not say how the company will fare against larger vendors.

This article is FinanceInTime's editorial view for information only. It is not investment, tax, legal or other advice, and it does not take your circumstances into account. Forecasts are uncertain and past results do not guarantee future ones.