Mintive
AI-agent infrastructure for governing and securing identities and permissions of AI agents.
Website: https://www.mintive.com/
Cover Block
From the public record
| Field | Value |
|---|---|
| Company | Mintive |
| Tagline | AI-agent infrastructure for governing and securing identities and permissions of AI agents |
| Headquarters | San Francisco/San Mateo, California, US [mintive.com, retrieved 2026] |
| Founded | 2023 [TechCrunch, October 2026] |
| Stage | Seed [TechCrunch, October 2026] |
| Business Model | SaaS |
| Industry | Security |
| Technology | AI / Machine Learning |
| Geography | North America |
| Growth Profile | Venture Scale |
| Founding Team | Co-Founders (2), Jiquan Ngiam and Vijay Vasudevan [TechCrunch, October 2026] |
| Funding Label | Seed |
| Total Disclosed Funding | ~$11,000,000 [TechCrunch, October 2026] |
Links
From the public record
- Website: https://www.mintive.com/
- Hiring: https://www.mintive.com/hiring
- Jobs: https://jobs.ashbyhq.com/Mintive
The Short Version
PUBLIC Mintive builds security and governance infrastructure for AI agents, a timely wedge as enterprises move from pilot deployments to production systems that need explicit controls around identity, permissions, and tool access [TechCrunch, October 2026] [mintive.com] [mintive.com, retrieved 2026]. The company was founded in 2023 and launched its product in February 2026, reaching more than 50 enterprise customers according to TechCrunch within months of launch, which is the clearest public signal that the category is moving from architecture discussion into budgeted deployment [TechCrunch, October 2026].
The product, originally branded MintMCP and now presented under the Mintive name, sits at the control layer for enterprise agents: it assigns agent identities, governs which tools and models those agents can call, and logs activity for oversight and auditability [mintive.com/about, retrieved 2026] [mintive.com/docs/intro, retrieved 2026] [cybercompanyprofiles.com, retrieved 2026]. That positioning matters because the differentiation appears to rest less on building another agent application and more on becoming the policy and access fabric around many agent workflows, with adjacent features such as monitoring, gateway controls, and budget management expanding the surface area over time [mintive.com/blog/meet-mintive, retrieved 2026] [cybercompanyprofiles.com, retrieved 2026].
The founding team, Jiquan Ngiam and Vijay Vasudevan, is presented as technically deep, and public materials tie the team to Google Brain, Waymo, and elite academic networks including Stanford, Berkeley, and CMU, although not every background detail is independently assigned founder by founder in the available reporting [TechCrunch, October 2026] [jobs.ashbyhq.com/mintive, retrieved 2026]. For investors, that profile is relevant because Mintive is selling into a technically demanding control plane problem where product credibility with engineering and security buyers tends to matter early [mintive.com/about, retrieved 2026].
On financing, Mintive announced a $7 million seed round led by Cowboy Ventures in October 2026, and TechCrunch reported $11 million in total funding; named backers include Coatue, Operator Collective, Hustle Fund, Maven Ventures, and several prominent angel investors [TechCrunch, October 2026] [mintive.com/about, retrieved 2026]. The business model is SaaS with custom pricing, which is consistent with an enterprise security product that likely sells through scoped deployments rather than self-serve motion [mintive.com/pricing, retrieved 2026].
Over the next 12 to 18 months, the key questions are whether Mintive can convert early customer count into durable platform usage, whether governance remains a distinct budget line rather than collapsing into broader AI infrastructure stacks, and whether company-reported customer logos and deployment claims continue to show independent corroboration in the market [TechCrunch, October 2026] [mintive.com/about, retrieved 2026] [jobs.ashbyhq.com/mintive, retrieved 2026]. The hiring mix across sales, customer success, marketing, and engineering suggests the company is trying to build both go-to-market capacity and implementation depth at the same time, which fits the stage but also raises the usual execution burden for a young enterprise software vendor [mintive.com/hiring, retrieved 2026] [jobs.ashbyhq.com, retrieved 2026].
Single-source, plausible -- Core company, founder, and funding facts are corroborated by TechCrunch and company materials, but several product and customer details remain company-sourced or partially corroborated.
Taxonomy Snapshot
| Axis | Value |
|---|---|
| Stage | Seed |
| Business Model | SaaS |
| Industry / Vertical | Security |
| Technology Type | AI / Machine Learning |
| Geography | North America |
| Growth Profile | Venture Scale |
| Founding Team | Co-Founders (2) |
| Funding | Seed, total disclosed approximately $11,000,000 |
The Company in Brief
PUBLIC
Mintive is an enterprise security software startup focused on the control layer around AI agents, specifically who those agents are, what systems they can access, and how their actions are logged and governed [mintive.com]. The company was founded in 2023 and is based in the San Francisco and San Mateo area, according to its public materials and hiring pages [mintive.com/about, retrieved 2026] [mintive.com/hiring, retrieved 2026].
The chronology is still short, but the public record is clear on a few milestones. Mintive launched its product in February 2026, according to TechCrunch, after forming around the problem of identity and permissions for production AI agents [TechCrunch, October 2026]. By October 2026, TechCrunch reported that the company had raised a $7 million seed round led by Cowboy Ventures and $11 million in total funding, while the company website separately names investors including Coatue, Maven Ventures, and Hustle Fund [TechCrunch, October 2026] [mintive.com/about, retrieved 2026].
The founder record is strongest at the company level and thinner on the legal-entity side. TechCrunch identifies Jiquan Ngiam and Vijay Vasudevan as co-founders, and Mintive's own site presents the company as infrastructure for the "agentic enterprise" under the Mintive brand, formerly MintMCP [TechCrunch, October 2026] [mintive.com, retrieved 2026]. No state filing or other public corporate registration detail was provided in the cited materials used for this section, so the overview rests on the operating brand, headquarters footprint, founding year, launch timing, and disclosed financing history [mintive.com/about, retrieved 2026] [TechCrunch, October 2026].
Single-source, plausible -- Founded date, headquarters footprint, and company description are supported by Mintive public materials; launch timing, co-founders, and funding details are supported by TechCrunch, but no state filing was provided for independent legal-entity confirmation.
What They Have Built
MIXED Product scope and deployment model
Mintive is pitching at a narrow but increasingly relevant control point: the layer that decides what an AI agent is allowed to access, which models or tools it can call, and how those actions are logged once agents move from experiment to production. Public materials describe the core product, originally branded MintMCP and now Mintive, as infrastructure for governing agent identity, permissions, and access, with a gateway model that sits between agents and enterprise systems [mintive.com, retrieved 2026] [mintive.com, retrieved 2026] [TechCrunch, October 2026]. The company says the product launched in February 2026, which matters because most of the operating history visible in public sources is measured in months rather than years [TechCrunch, October 2026].
The public feature set is framed around enterprise controls rather than model performance. Mintive says customers can run AI agents in production with controls over reachable tools, model calls, and spend, alongside centralized observability and auditability [mintive.com, retrieved 2026] [mintive.com, retrieved 2026]. Its site also describes support for managed SaaS, VPC, and self-hosted deployment options, delegated authentication through enterprise identity providers including Okta, Google Workspace, and Microsoft Entra ID, plus SIEM export, encryption in transit and at rest, uptime SLAs, and data residency options [mintive.com, retrieved 2026] [mintive.com, retrieved 2026] [mintive.com, retrieved 2026]. Those claims, taken together, suggest Mintive wants to be evaluated less as a developer utility and more as enterprise control-plane software for agent traffic.
MIXED Product surfaces and technical posture
The architecture visible in public sources appears to have several product surfaces around that control plane. Mintive's documentation and blog describe an MCP gateway, an LLM gateway, Mint Guard for screening tool traffic, Agent Monitor for visibility into agent activity, and beta "Coworker Agents" in Slack, while third-party coverage describes the core offer as a gateway that controls which tools agents can call and logs every call [mintive.com, retrieved 2026] [mintive.com, retrieved 2026] [cybercompanyprofiles.com, retrieved 2026]. The company also says it can host MCP connectors, wrap local servers with OAuth, and apply custom policy logic that calls external DLP services or business rules, which points to a product strategy built around adapting fast-moving open agent tooling to enterprise security requirements [mintive.com, retrieved 2026] [mintive.com, retrieved 2026] [mintive.com, retrieved 2026].
There are early signs of technical credibility, but the evidence is uneven and mostly company-authored. Mintive cites use cases such as natural-language access to Oracle systems, remote servers connecting tools like Glean and Guru, and customer deployments at firms including Ro, Coursera, Modern Treasury, and Stability AI [mintive.com, retrieved 2026] [mintive.com, retrieved 2026] [mintive.com, retrieved 2026] [TechCrunch, October 2026]. A separate reviewer wrote that a basic GitHub setup worked end to end, while SSO and RBAC sat behind the enterprise tier, which is directionally consistent with the company's positioning around governance-heavy deployments [manveerc.substack.com, retrieved 2026]. The technical stack itself is not publicly detailed; the only safe inference from hiring is that Mintive is still building out engineering, sales engineering, customer success, and go-to-market capacity around the current platform, rather than advertising a fully fixed product footprint (inferred from job postings) [jobs.ashbyhq.com, retrieved 2026] [mintive.com, retrieved 2026].
Unconfirmed -- Material product and security claims in this section rely primarily on Mintive's website, docs, and blog, with limited third-party corroboration from TechCrunch and one external product review.
Market Size and Demand
PUBLIC
The market matters now because enterprise AI agents are moving from experiment to production, and that shift turns identity, access control, and auditability from technical nice-to-haves into budgeted security requirements [TechCrunch, October 2026] [mintive.com, retrieved 2026].
A clean TAM build is not possible from the supplied evidence because no named third-party market report for AI agent governance, agent identity, or MCP gateways is included in the research pack. The public record is still enough to frame the category: Mintive sells controls for how AI agents authenticate, what tools they can call, which models they invoke, and how those actions are logged in production environments [mintive.com, retrieved 2026] [mintive.com, retrieved 2026]. That places the company at the intersection of several larger adjacent budgets rather than in a mature standalone line item, namely identity and access management, API and gateway infrastructure, cloud security, and emerging AI governance software [TechCrunch, October 2026] [mintive.com, retrieved 2026].
The best public read on demand is bottom-up rather than top-down. TechCrunch reported in October 2026 that Mintive launched in February 2026 and had reached more than 50 enterprise customers within months, while describing the company as focused on governing and securing AI-agent identities and permissions [TechCrunch, October 2026]. That does not size the market, but it does indicate that early enterprise demand exists for control-plane software that sits between agents, enterprise identity systems, and enterprise tools.
The demand drivers are straightforward and visible in the product language itself. Mintive emphasizes delegated permissions, model-call controls, audit logs, SIEM export, enterprise authentication, and policy enforcement, which suggests customers are not buying generic experimentation tooling so much as risk controls for production deployments [mintive.com, retrieved 2026] [mintive.com, retrieved 2026] [mintive.com, retrieved 2026]. The cited customer set also skews toward organizations with meaningful data sensitivity or operational complexity, including Coursera, Modern Treasury, Stability AI, Ro, and the University of Michigan according to TechCrunch and the company site [TechCrunch, October 2026] [mintive.com, retrieved 2026].
Adjacent markets matter because buyers may solve this problem with nearby tools before they add a net-new platform. OneTrust comes from governance and compliance, Kong from API gateway and service control, and Portkey or LiteLLM from LLM traffic management, while homegrown IAM and logging stacks remain the default substitute for some engineering-led teams [cybercompanyprofiles.com, retrieved 2026]. The practical implication is that Mintive is not selling into an empty whitespace category. It is trying to consolidate pieces of identity, policy, observability, and gateway control that are already funded elsewhere in the stack [mintive.com, retrieved 2026] [mintive.com, retrieved 2026].
Regulatory and macro forces appear supportive, even if the evidence here is mostly indirect. Mintive's public materials repeatedly stress audited security programs, SOC 2 Type II status, HIPAA alignment, SAML or OIDC support, penetration testing, data residency options, and export into enterprise security systems, which is usually how vendors respond to procurement pressure rather than how they market to hobbyist users [mintive.com, retrieved 2026] [mintive.com, retrieved 2026] [mintive.com, retrieved 2026]. In other words, the category's growth is likely tied less to consumer AI enthusiasm and more to enterprise requirements around internal controls, traceability, and least-privilege access as agents begin to touch financial, ERP, and knowledge systems [mintive.com, retrieved 2026] [mintive.com, retrieved 2026].
| Cited market signal | Figure | Source |
|---|---|---|
| Product launch timing | February 2026 | [TechCrunch, October 2026] |
| Enterprise customers | 50+ | [TechCrunch, October 2026] |
| Total funding raised | $11M | [TechCrunch, October 2026] |
| Confirmed seed round | $7M | [TechCrunch, October 2026] |
The table is not a market-sizing model, but it does show the core public signal: investor capital and early customer adoption are arriving before a standardized category taxonomy has fully formed. That usually describes an emerging control layer, not a saturated application market.
Unconfirmed -- This section relies on one independent article for the core adoption and funding signals, while most market-shaping evidence comes from Mintive's own product materials rather than third-party market reports.
Who Else Is Fighting for This
MIXED Mintive is positioning itself between traditional identity-and-governance vendors and the newer crop of AI gateway tooling, with a narrower claim: govern what AI agents are allowed to do, which tools and models they can call, and how those actions are audited [TechCrunch, October 2026] [mintive.com] [cybercompanyprofiles.com].
| Company | Positioning | Stage / Funding | Notable Differentiator | Source |
|---|---|---|---|---|
| Mintive | AI-agent infrastructure focused on identity, permissions, gateway control, and governance for enterprise AI agents | Seed, $11M total raised; $7M seed announced October 2026 | Purpose-built around agent identity plus control of tool calls and model calls, with enterprise governance features such as auditability and policy enforcement | [TechCrunch, October 2026] [mintive.com] |
| OneTrust | Enterprise governance and compliance platform, broadly associated with privacy, risk, and data governance | Not included in provided research | Incumbent enterprise governance footprint and established compliance relationships | [Structured facts] |
| Kong | API and gateway infrastructure vendor | Not included in provided research | Strong gateway heritage, which matters if enterprises treat agent traffic as an extension of API traffic | [Structured facts] |
| Portkey | LLM gateway and observability tooling | Not included in provided research | Closer to model routing and LLM operations than identity-first governance | [Structured facts] |
| LiteLLM | Developer-facing layer for managing model access and calls | Not included in provided research | Open, developer-friendly abstraction across model providers | [Structured facts] |
| TrueFoundry | Platform for deploying and operating AI systems | Not included in provided research | Broader AI infrastructure posture, potentially attractive where buyers want one control plane | [Structured facts] |
The field breaks into three practical groups.
Mintive's clearest edge today appears to be focus. Public materials consistently frame the company around agent identity, delegated permissions, gateway controls, and auditability, rather than around generic LLM orchestration or app-layer copilots [mintive.com] [mintive.com/docs/intro] [TechCrunch, October 2026]. That specialization is helped by founder-market fit and early enterprise logos, including Ro, Coursera, Modern Treasury, and Stability AI, according to TechCrunch and Mintive's own site [TechCrunch, October 2026] [mintive.com/about]. Still, this edge looks perishable unless Mintive turns product depth into workflow lock-in, because model gateways, API gateways, and broader enterprise AI platforms can all move toward the same control surface over time [mintive.com]
The exposure is also fairly plain. Mintive also does not appear, from the public record provided here, to own a proprietary distribution channel beyond founder network, investors, and early customer references, which makes go-to-market execution a material variable rather than a solved one [TechCrunch, October 2026] [mintive.com/about].
The most plausible 18-month scenario is a market that separates into specialist control layers and broader platform bundles, with buying motion determined by who owns the deployment path. Mintive is a winner if enterprises increasingly distinguish agent identity and tool-permissioning from general LLM observability, because that framing favors a purpose-built vendor with narrow messaging and early customer proof points [TechCrunch, October 2026] [mintive.com/about].
Opportunity
PUBLIC The prize here is not another point security tool, but a control plane that could sit in the critical path of how enterprises deploy, authenticate, observe, and constrain AI agents across internal systems [TechCrunch, October 2026] [mintive.com, retrieved 2026].
The headline opportunity is straightforward. If AI agents become a durable enterprise software layer, someone has to manage their identities, permissions, tool access, model calls, and audit trails, and Mintive is trying to own that layer rather than any single agent application [TechCrunch, October 2026] [mintive.com, retrieved 2026]. That outcome looks reachable, not merely rhetorical, because the company launched in February 2026 and was reported at 50+ enterprise customers by October 2026, with named users including Ro, Modern Treasury, Coursera, and Stability AI across both independent coverage and company materials [TechCrunch, October 2026] [mintive.com, retrieved 2026].
The public record also suggests Mintive is selling into a real enterprise pain point. Its product claims center on delegated permissions, tool-call governance, model-call controls, SIEM export, enterprise authentication, and deployment options including managed SaaS, VPC, and self-hosted, which is the kind of feature set large security and infrastructure teams usually ask for when a new software surface starts touching production systems [mintive.com, retrieved 2026] [mintive.com, retrieved 2026]. The investor roster matters here as well: Cowboy Ventures led the $7 million seed, while Coatue, Operator Collective, Hustle Fund, Maven Ventures, Frederic Kerrest, Jeff Dean, Andrej Karpathy, Michael Grinich, Scott Belsky, Tom Willerer, and Brian Shultz were all publicly associated with the company, which raises the odds that Mintive gets early access to design partners, talent, and distribution channels that a younger infrastructure vendor normally lacks [TechCrunch, October 2026] [mintive.com, retrieved 2026].
| Scenario | What happens | Catalyst | Why it's plausible |
|---|---|---|---|
| Security control plane for enterprise agents | Mintive becomes the default governance layer that enterprises standardize on before allowing internal or customer-facing agents into production | Category shift from pilot agents to production deployments, where identity, permissioning, logging, and policy controls move from optional to required [mintive.com, retrieved 2026] | Mintive already frames the product around agent identity and governance, and TechCrunch reported 50+ enterprise customers less than a year after launch [TechCrunch, October 2026] [mintive.com, retrieved 2026] |
| Gateway layer across MCP and LLM traffic | Mintive wins by owning both tool-call and model-call routing, giving enterprises one place to enforce policy, budgets, and observability across heterogeneous agent stacks | Broader adoption of MCP servers and multi-model agent architectures, which increases operational complexity and makes a unified gateway more valuable [mintive.com, retrieved 2026] [mintive.com, retrieved 2026] | The company publicly positions itself as both an MCP and LLM gateway, with centralized observability, model-call budgets in early access, and one-click deployment for governed services [mintive.com, retrieved 2026] [cybercompanyprofiles.com, retrieved 2026] |
| Embedded governance inside agent ecosystems | Mintive becomes the enterprise governance layer used by adjacent agent builders, coding tools, and connector ecosystems rather than selling only direct top-down security software | A category-tipping integration or partner motion, such as the reported Cursor hooks partnership for governing coding agents [LinkedIn] | Public evidence already points to ecosystem behavior, including a reported Cursor partnership and a registry claim of 10,000+ MCP servers, which, if adopted by customers, could increase switching costs over time [LinkedIn] [mintive.com, retrieved 2026] |
The compounding dynamic is visible, even if still early. Each new enterprise deployment can deepen Mintive's position in three ways: more governed connectors and servers inside the customer environment, more policy logic tied to internal identity providers such as Okta or Microsoft Entra, and more monitoring or audit workflows exported into existing security systems [mintive.com, retrieved 2026] [mintive.com, retrieved 2026]. Once a platform is sitting between agents and the systems they can reach, replacement becomes less about swapping software and more about rebuilding permissions, observability, and compliance workflows.
There is a second flywheel if the product surface broadens as advertised. Mintive says it spans MCP gateway, LLM gateway, Mint Guard, Agent Monitoring, and Coworker Agents, while also hosting connectors and offering managed SaaS, VPC, and self-hosted deployment options [mintive.com, retrieved 2026] [mintive.com, retrieved 2026]. If those modules land in the same accounts, the company can move from a narrow security wedge to a broader platform position, where one initial governance purchase expands into deployment, hosting, monitoring, and policy enforcement. The hiring pattern supports that expansion thesis: open roles span sales, growth, revops, customer success, and engineering, which is consistent with a company building a repeatable go-to-market motion rather than serving only a handful of bespoke design partners [mintive.com, retrieved 2026] [jobs.ashbyhq.com, retrieved 2026].
The size of the win is best framed through scenario analysis because no reliable public market-sizing figure was provided in the available sources. In the strongest case, Mintive could become part of the core enterprise security and infrastructure stack for AI agents, adjacent to identity, API management, and governance vendors such as OneTrust, Kong, and newer agent infrastructure players named in the competitive set [TechCrunch, October 2026] [cybercompanyprofiles.com, retrieved 2026]. If the "security control plane for enterprise agents" scenario plays out, the company could plausibly grow into a multibillion-dollar outcome (scenario, not a forecast), because it would be selling a high-importance control layer into large enterprises with room for platform expansion across authentication, gatewaying, observability, and policy management [mintive.com, retrieved 2026] [TechCrunch, October 2026]. The public evidence does not yet support a tighter valuation analog than that, but it does support the underlying reason the upside exists: Mintive is aiming at a control point that becomes more valuable as enterprises let more software act on their behalf.
Single-source, plausible -- Core upside claims rely on one independent publisher, TechCrunch, supplemented by company materials and LinkedIn posts for product breadth and ecosystem signals.
Sources
From the public record
[mintive.com, retrieved 2026] Mintive | MCP & LLM Gateway for the Agentic Enterprise | https://www.mintive.com/
[TechCrunch, October 2026] Agent identity startup Mintive lands Cowboy Ventures, Coatue, angels Jeff Dean, Andrej Karpathy | https://techcrunch.com/2026/10/29/agent-identity-startup-mintive-lands-cowboy-ventures-coatue-angels-jeff-dean-andrej-karpathy/
[mintive.com/about, retrieved 2026] About Mintive: Infrastructure for the agentic enterprise | https://www.mintive.com/about
[mintive.com/docs/intro, retrieved 2026] Introduction to Mintive | Mintive | https://www.mintive.com/docs/intro
[cybercompanyprofiles.com, retrieved 2026] Mintive: Funding, Competitors and Strategy | https://cybercompanyprofiles.com/companies/mintive
[jobs.ashbyhq.com/mintive, retrieved 2026] Mintive hiring page | https://jobs.ashbyhq.com/mintive/dc7a3f82-153d-46c4-b7eb-d65cbb3ed33f
[mintive.com/pricing, retrieved 2026] Pricing | Mintive - Request Your Custom Quote | https://www.mintive.com/pricing
[mintive.com/hiring, retrieved 2026] Mintive hiring page | https://www.mintive.com/hiring
[jobs.ashbyhq.com, retrieved 2026] Mintive hiring page | https://jobs.ashbyhq.com/mintive/dc7a3f82-153d-46c4-b7eb-d65cbb3ed33f
[mintive.com/blog/meet-mintive, retrieved 2026] Meet Mintive | Mintive Blog | https://www.mintive.com/blog/meet-mintive
[manveerc.substack.com, retrieved 2026] Best MCP Gateways | https://manveerc.substack.com/p/best-mcp-gateways
[LinkedIn] Jeremy Stribling - Software Engineer building Cursor | LinkedIn | https://www.linkedin.com/in/jeremy-stribling-94b60455/
Articles about Mintive
- Mintive Lands 50 Enterprise Customers for Its AI Agent Identity Gateway — The startup, backed by Cowboy Ventures and angels like Jeff Dean, is selling identity infrastructure for AI agents to companies like Coursera and Modern Treasury.