Mintive builds infrastructure for the agentic enterprise

Why we started Mintive

We started Mintive (formerly MintMCP) after watching the first enterprise agents reach production. They had broad access to company systems and left no clear record of which agent took which action.

So we treat each agent the way a company treats an employee. It gets its own identity, only the access its work requires, and a record of what it did. Most people carry far more permissions than any one task needs, and an agent that borrows a person's token inherits all of them.

Mintive's four founders

Backed by

We've raised $11 million, most recently in a round led by Cowboy Ventures. Coatue, Operator Collective, Hustle Fund, Burst Capital, and Maven Ventures also back us, along with Okta co-founder Frederic Kerrest, WorkOS founder and CEO Michael Grinich, former Adobe Chief Product Officer Scott Belsky, Discovery Loop co-founder and former Google Chief Scientist Jeff Dean, and OpenAI co-founder Andrej Karpathy.

“We're early in the enterprise adoption of AI with the major gating factors being governance, IT and security team ability to oversee, identity management, and security. Mintive offers the critical, scalable, trusted third party scaffolding businesses need - to deliver agent identity, permissions and tracing, regardless of the underlying models.”

Aileen Lee, Founder and Managing Partner, Cowboy Ventures
Cowboy Ventures
Coatue Management
Operator Collective
Hustle Fund
Burst Capital
Maven Ventures

Trusted by teams putting agents to work

More than 50 customers run 2.5 million agent calls a day through Mintive, including Ro, Modern Treasury, Braze, Coursera, the University of Michigan, Apax, and Stability AI. Read customer stories.

Scaling the AI pyramid

We think about agent adoption in four levels. Most companies are still near the bottom, and moving up depends on having scoped data access, audit logs, and governance in place first.

Mintive is that foundation. Agents reach company tools through MCP, each with its own identity and permissions scoped to the task. That moves a company from level 1 to level 2 with controls in place.

Monitoring shows what agents do in production, including the scheduled work that runs overnight. Coworker agents work in Slack and Teams alongside your people, keep long-term memory, and take on work before anyone asks.

Further up, the way companies organize their work and software will change. We are building toward that from trusted access up.

1

Chat

Tools like ChatGPT and Claude help people write, analyze, and think. The AI advises, but it doesn't own any of the work yet. Most people are here.

2

Co-work

The AI starts doing bounded tasks: editing files, updating spreadsheets, writing docs, and changing tickets. It moves from giving advice to finishing tasks, and a lot of companies are adopting this right now.

3

Co-worker

Many people work with the same agents at once. The agents show up in Slack, keep memory across time, and can tell you what they did yesterday. They start to feel like teammates.

4

The machine

Agents run entire workflows. More company software is built around agent-managed systems.

What Mintive does for agent governance

Every agent needs models, tools, and company data. The MCP gateway governs tools and the LLM gateway governs models, so one set of rules covers Claude, ChatGPT, Copilot, Cursor, and the agents you build.

Agent identities

Every agent gets its own identity, credentials, and least-privilege scope. Revoke one agent without touching anyone else.

MCP gateway

Route every agent and MCP tool through one gateway you control. Approve the servers your team can use, bundle them by role, and switch off destructive tools.

LLM gateway

Route model calls from Claude Code, Codex, and your agents through one gateway. Give each team its own budget and allowed models, and switch between closed and open models, from providers like OpenAI, Anthropic, Fireworks, and OpenRouter, without rewiring agents.

Monitoring & cost

Every tool call traces back to a person, agent, server, and model, with spend and tamper-proof audit logs you can export to your SIEM.

Security & Mint Guard

Mint Guard screens calls for prompt injection, exposed secrets, and personal data. Run your DLP tools inline and reach systems inside your network over a private tunnel.

Coworker agents

Agents work alongside your team in Slack and Teams, with long-term memory and schedules, on the same governed access.

Founded by

A team of scientists and builders from Stanford, Google Brain, Coursera, Applied Intuition, Snorkel, and Amazon. Come work with us.

Jiquan Ngiam
Jiquan Ngiam
Google, Coursera
Bozhi See
Bozhi See
Brilliant Home, Shopkick
Phillip Bensaid
Phillip Bensaid
Disney+, Amazon
SOC 2
HIPAA attestation

Secure & compliant

Mintive is SOC 2 Type II audited, with continuous compliance monitoring via Drata. Enterprise SSO, complete audit trails, PII detection, and role-based access control are built into every layer of the platform. EU and US data residency options are available. Customers handling protected health information can request our HIPAA documentation. We sign BAAs.

Visit our Trust Center or contact security@mintive.com for compliance documentation.

Ready to put your agents to work safely?