Mintive: AI Agent Identity and Enterprise Governance
Mintive builds enterprise infrastructure that helps organisations deploy and manage AI agents securely. Its platform provides agent identities, controlled access to business tools, model gateways, and activity monitoring. On October 29, 2026, Mintive announced $11 million in total funding, including a new $7 million round led by Cowboy Ventures, to expand infrastructure for the agentic enterprise.
Visit Official Website ↗Mintive at a Glance: Company Overview
Mintive is an enterprise AI infrastructure startup that helps businesses manage the identity, permissions, and activity of AI agents. As employees increasingly use AI systems to interact with company applications and data, organisations need a way to control which tools those agents can access and which actions they can perform.
Previously known as MintMCP, Mintive provides an infrastructure layer for enterprise AI agents. Its products include an MCP gateway for governing tool access, an LLM gateway for controlling model usage, agent identity management, activity monitoring, and AI coworkers that can operate alongside employees in Slack and Microsoft Teams.
Mintive’s approach treats an AI agent as a distinct identity rather than allowing it to inherit all the permissions of the employee who configured it. This helps organisations establish more specific access rules and trace actions back to the person, agent, model, and tool involved.
| Company Name | Mintive |
| Previous Name | MintMCP |
| Industry | Enterprise AI infrastructure, AI agent security, identity and access management |
| Founders | Jiquan Ngiam and Vijay Vasudevan |
| CEO | Jiquan Ngiam |
| CTO | Vijay Vasudevan |
| Product Launch | Publicly launched in 2026; the company said its product launched in February 2026 |
| Headquarters | Not independently confirmed in the cited company announcement |
| Primary Focus | AI agent identity, access governance, model routing, and activity monitoring |
| Key Products | MCP Gateway, LLM Gateway, Agent Identity, Agent Monitor, Coworker Agents, and Mint Guard |
| Latest Funding Round | $7 million round announced October 29, 2026 |
| Total Disclosed Funding | $11 million, according to the company’s announcement |
| Lead Investor | Cowboy Ventures |
| Other Investors | Operator Collective, Burst Capital, Coatue, Hustle Fund, and Maven Ventures |
| Notable Angel Investors | Frederic Kerrest, Michael Grinich, Scott Belsky, Jeff Dean, and Andrej Karpathy, among others |
| Valuation | Not publicly disclosed |
| Official Website | mintive.com |
What Does Mintive Do?
Mintive provides the infrastructure organisations need to govern AI agents that interact with internal applications, business data, and enterprise tools. Its platform addresses a key challenge in enterprise AI adoption: agents need enough access to complete tasks, but they should not automatically receive unrestricted access to the systems used by the employees they serve.
Governs AI Agent Access
Mintive gives AI agents distinct identities and scoped credentials. Organisations can control which tools an agent may use and revoke an agent’s access without having to change the identity of the employee it serves.
Controls AI Model Usage
Its LLM Gateway routes model requests through a central control layer. Teams can define approved models, apply policies across providers, and manage usage budgets for different groups.
Monitors Agent Activity
Mintive records activity across agents, models, and tools. Its monitoring capabilities help security and IT teams investigate which person initiated a task, which agent acted, and which systems were accessed.
Mintive Products and Key Capabilities
Mintive offers several connected products designed to work as a governance layer for enterprise AI. The company aims to let organisations use different AI models and agents while maintaining consistent policies for access, monitoring, and security.
1. MCP Gateway
The Model Context Protocol (MCP) Gateway controls how AI agents connect to tools and enterprise services. Administrators can approve MCP servers, organise tools into role-based bundles, and restrict access to specific actions.
2. LLM Gateway
The LLM Gateway centralises model access and routing. It allows organisations to set model policies, manage budgets by group, and switch between supported model providers without rebuilding each agent integration.
3. Agent Identity
Agent Identity gives each AI agent its own identity and scoped, revocable credentials. The aim is to prevent agents from automatically inheriting all the permissions associated with a human user’s account.
4. Agent Monitor
Agent Monitor traces agent activity across people, models, servers, and tools. The company says administrators can review tool calls, timestamps, arguments, results, and usage costs to understand how agents operate.
5. Mint Guard
Mint Guard is Mintive’s security layer for screening AI interactions. The company describes capabilities for detecting prompt injection, exposed secrets, and personal data before requests reach enterprise systems.
6. Coworker Agents
Mintive’s coworker agents operate alongside employees in Slack and Microsoft Teams. They can use long-term memory and scheduled work while remaining subject to the identities, permissions, and monitoring configured through the platform.
How Mintive Secures Enterprise AI Agents
Enterprise AI agents may need to read documents, query internal systems, update records, or trigger business processes. Giving these agents broad access creates security and governance challenges. Mintive’s architecture aims to separate agent identity, tool access, model access, and activity records from the underlying AI model.
Step 1: Assign an Agent Identity
Each agent receives a distinct identity and credentials. This makes it possible to manage the agent separately from the employee who requested or configured it.
Step 2: Define Tool Permissions
Administrators decide which tools, services, and actions the agent can access. Access can be limited to the requirements of a particular role or task.
Step 3: Apply Model Policies
Model requests can pass through the LLM Gateway, where organisations manage approved providers, usage policies, and spending limits.
Step 4: Screen Requests
Mint Guard is designed to identify risks such as prompt injection, exposed secrets, and personal data in AI interactions before they reach connected company systems.
Step 5: Record Agent Actions
Monitoring and audit logs help teams trace interactions among people, agents, models, and tools. This information can support troubleshooting, security investigations, and cost reviews.
Step 6: Review and Revoke Access
Administrators can review activity and revoke an agent’s credentials when access is no longer required. Organisations should still define appropriate approval processes for sensitive operations.
Why Do Enterprises Need AI Agent Identity and Governance?
Traditional identity and access management systems are designed primarily around human users, applications, and established service accounts. AI agents introduce a different challenge because they can use multiple tools, interact with different models, and execute multi-step tasks on behalf of people or teams.
When an agent uses a human employee’s credentials, it may inherit access that is broader than the task requires. When several agents share credentials, it can also become difficult to determine which agent performed a particular action. Mintive aims to address these problems with distinct agent identities, controlled tool access, and detailed activity records.
Least-Privilege Access
Give each agent access to only the tools and data required for its task, rather than allowing it to inherit an employee’s full set of permissions.
Centralised Governance
Apply consistent policies across different AI agents, model providers, and connected enterprise tools instead of configuring every connection independently.
Auditability and Visibility
Build a record of how agents interact with enterprise systems so teams can investigate errors, review activity, and understand operational costs.
Mintive Enterprise Use Cases
Mintive is designed for organisations that want to move AI agents from experiments into production while maintaining control over access to company data and systems. Its infrastructure can support use cases across IT, security, finance, operations, and employee productivity.
IT and Enterprise Systems
- Govern access to internal tools and applications.
- Control AI connections to MCP servers.
- Manage permissions for agent-driven workflows.
- Trace tool calls and system interactions.
Cybersecurity Teams
- Monitor agent activity across enterprise systems.
- Investigate potentially unsafe tool interactions.
- Screen for prompt injection and exposed secrets.
- Review access and audit records.
Finance and Business Operations
- Control agents that interact with financial tools.
- Manage access to internal business records.
- Apply approved model and tool policies.
- Track usage costs across teams.
Software Engineering
- Govern tools used by coding agents.
- Control model access for development workflows.
- Review agent activity across connected systems.
- Apply limits to potentially destructive tools.
Employee Productivity
- Deploy agents inside Slack and Microsoft Teams.
- Connect agents to approved company information.
- Support scheduled and proactive work.
- Keep agent permissions under enterprise control.
Education and Healthcare
- Govern AI access to approved organisational services.
- Apply access restrictions to sensitive information.
- Review audit trails for agent activity.
- Evaluate data residency and compliance requirements.
Mintive Customers and Enterprise Adoption
Mintive reported more than 50 paying customers in its October 2026 funding announcement. The company named organisations including Ro, Modern Treasury, Braze, Coursera, the University of Michigan, Apax Partners, and Stability AI among its customers.
These examples illustrate the types of environments where agent governance may be important, from financial technology and healthcare to education and AI development. The specific products deployed and the scope of each customer relationship can differ.
Modern Treasury
Mintive describes a deployment where an internal agent searches authorised tools and information sources to help technical account managers investigate customer cases. The company says the deployment went live in under three weeks.
University of Michigan
Mintive says its MCP Gateway is used to govern access to approved tools in a university environment serving a community of approximately 150,000 people.
Ro and Stability AI
Mintive lists Ro and Stability AI among its customers. These organisations operate in environments where controlling access to information and maintaining visibility into AI activity can be important.
Mintive Funding History and Investors
On October 29, 2026, Mintive announced $11 million in total funding, including a new $7 million round led by Cowboy Ventures. The company said the round would support its infrastructure for enterprise AI agents. The announcement did not disclose a company valuation.
The company says its product launched in February 2026. It was previously known as MintMCP.
Led by Cowboy Ventures, with participation from Operator Collective and Burst Capital, alongside existing investors.
Total funding reported by the company after its latest announcement. Valuation was not disclosed.
Other investors named in connection with the funding include Coatue, Hustle Fund, and Maven Ventures. Angel investors include Frederic Kerrest, co-founder of Okta; Michael Grinich, founder of WorkOS; Scott Belsky, former Chief Product Officer at Adobe; Jeff Dean, former Google Chief Scientist; and Andrej Karpathy, co-founder of OpenAI.
Mintive vs. Traditional Identity and AI Management Tools
Mintive operates in the emerging market for AI agent management and governance. Traditional identity management tools focus on people and established applications, while AI development platforms help teams build agents. Mintive focuses on the layer that governs agent identities, model usage, tool permissions, and activity across those systems.
| Area | Traditional Approach | Mintive’s Approach |
|---|---|---|
| Agent Identity | Agents may rely on credentials associated with a human user or service account. | Gives agents distinct identities and scoped, revocable credentials. |
| Tool Access | Access may be configured separately across applications and integrations. | Uses an MCP Gateway to centralise control over approved tools and servers. |
| Model Access | Different applications may connect to models through separate configurations. | Uses an LLM Gateway to route model requests and apply policies across providers. |
| Activity Monitoring | Logs may be spread across models, applications, and integration services. | Agent Monitor aims to connect activity across people, agents, models, and tools. |
| Security Screening | Security controls may depend on individual applications and existing security products. | Mint Guard is designed to screen AI interactions for prompt injection, exposed secrets, and personal data. |
| Enterprise AI Deployment | Teams may need to configure permissions and monitoring for each agent implementation. | Provides a shared governance layer for agents using supported models and enterprise tools. |
These differences describe product positioning rather than a complete comparison of every vendor. Enterprises should evaluate integrations, supported identity providers, security certifications, audit requirements, performance, and total cost before selecting a platform.
Security, Privacy, and Compliance
AI agent governance involves more than deciding which model an organisation can use. Agents may access sensitive documents, execute tool calls, and transfer information between applications. Businesses need controls over agent credentials, access permissions, data handling, and records of activity.
Access Control
Mintive describes role-based access controls and scoped agent credentials designed to restrict access to the tools and resources required for a task.
Audit Trails
The platform records agent activity and related tool calls, helping teams understand what happened during an AI-driven workflow and investigate unexpected behaviour.
Data Protection
Mintive says Mint Guard screens for prompt injection, exposed secrets, and personal data. The company also describes EU and US data residency options and SOC 2 Type II auditing.
Organisations evaluating Mintive should review the current security documentation, scope of its SOC 2 Type II audit, data processing terms, retention policies, and any relevant healthcare or privacy requirements directly with the company. Security controls reduce risk but do not remove the need for governance and oversight.
Frequently Asked Questions About Mintive
What is Mintive?
Mintive is an enterprise AI infrastructure company that helps organisations govern AI agents through agent identities, controlled tool access, model gateways, and activity monitoring. It was previously known as MintMCP.
How much funding has Mintive raised?
Mintive reported $11 million in total funding on October 29, 2026. This includes a newly announced $7 million round led by Cowboy Ventures.
Who invested in Mintive?
Cowboy Ventures led its latest $7 million round. Other investors named by the company include Operator Collective, Burst Capital, Coatue, Hustle Fund, and Maven Ventures, alongside angel investors.
Who founded Mintive?
Mintive was founded by Jiquan Ngiam and Vijay Vasudevan. Ngiam is the company’s CEO and was an early engineer at Coursera before working on neural networks at Google Brain. Vasudevan is the CTO and was one of the co-creators of TensorFlow while at Google Brain, according to company and press profiles.
What problem does Mintive solve?
Mintive addresses the challenge of managing AI agents that access company systems. It helps organisations assign distinct agent identities, restrict tool access, govern model usage, and trace agent actions.
What is an MCP Gateway?
An MCP Gateway manages how AI agents access tools exposed through the Model Context Protocol. Mintive’s gateway lets administrators control approved servers, group tools by role, and restrict access to particular actions.
What is Mintive’s LLM Gateway?
The LLM Gateway routes model requests through a central control layer. It helps enterprises apply model policies, manage spending budgets, and switch between supported AI providers.
How does Mintive manage AI agent identity?
Mintive assigns each agent a distinct identity and scoped, revocable credentials. This allows an organisation to limit an agent’s permissions and revoke its access separately from the employee it serves.
What is Mint Guard?
Mint Guard is Mintive’s security layer for screening AI interactions. The company describes capabilities for detecting prompt injection, exposed secrets, and personal data before requests reach company systems.
Which companies use Mintive?
Mintive lists customers including Ro, Modern Treasury, Braze, Coursera, the University of Michigan, Apax Partners, and Stability AI. The company reported more than 50 paying customers in October 2026.
Is Mintive an AI agent builder?
Mintive’s main focus is infrastructure for deploying and governing agents rather than only building AI models. It also offers coworker agents that operate alongside employees in Slack and Microsoft Teams.
What is Mintive’s valuation?
Mintive’s valuation was not publicly disclosed in its October 29, 2026 funding announcement.
What is Mintive’s official website?
Mintive’s official website is mintive.com, where organisations can explore its products, company information, and enterprise AI governance capabilities.