MintMCP. Founding Team · Inference Search Partners
Now Hiring · Founding Team

MintMCP

Founding Engineer & Founding GTM
Seed Stage Palo Alto, CA Backed by Karpathy & Jeff Dean

AI agents are everywhere in the enterprise. Nobody knows what they're doing.

When Claude, Cursor, or a custom agent touches Slack, Snowflake, or GitHub, who can see what happened? Right now: almost no one. There's no identity layer for agents, no permissions framework, no audit trail that works across all your AI systems at once.

MintMCP gives every user and every AI agent its own identity and permissions, so companies can see exactly which agent did what, enforce guardrails in real time, and govern every AI-to-data connection from a single control plane, regardless of which model or agentic framework is doing the work.

The Model Context Protocol (MCP) is the emerging standard for how agents connect to tools and data: Anthropic built it, and OpenAI, Microsoft, and Google all adopted it. MintMCP's bet is that the governance layer sitting on top of MCP will be the most critical piece of enterprise AI infrastructure built in the next five years, and they're building it now.

Stage
Seed
Location
Palo Alto, CA

Week-over-week revenue growth, almost entirely inbound.

Since launching in February, MintMCP has signed dozens of enterprise customers and is growing without a single outbound sales hire. Customers are finding us.

$$$
week-over-week rev growth
Almost entirely inbound, no outbound sales team yet
Dozens
Enterprise customers signed since February launch
Including Braze, Arlo, and Coursera
20+
Custom MCP servers at most advanced customers
Reflecting how each business actually operates

Fewer than 10 people, 25+ agents already on the team.

Those agents run real work across sales, revenue operations, and engineering, and the plan is to keep agentifying as much of the company as possible. If you want to build agents and use them every day as the core of how you work, you'll find few other roles like this.

🛠
You build it.
Founding team, no layers, no filtered specs: you make the architectural decisions that will still be running when this is a large company. The control plane, the governance engine, the telemetry, it's yours to design.
You use it.
The founders run real agent workflows on MintMCP every day, so the day customers get a feature, you feel it too. The build-use feedback loop is as tight as it gets.
📈
You sell it.
A direct line from the code you write to the customers using it. You'll talk to CTOs and platform leads, hear what they need, and go back and build it, with no abstraction layer between you and the market.
Building a 20-agent GTM team with Claude + MCPs
Jiquan building live, the clearest way to see how the founders think and how the product actually works.
Watch now

Two engineers who've built things that ended up running under most of the world's AI.

JN
Jiquan Ngiam
Co-Founder & CEO
  • Stanford CS, researched under Andrew Ng in the deep learning lab
  • Co-created Stanford's first online ML course, before Coursera was Coursera
  • Founding team at Coursera, scaled from zero to millions of learners globally
  • Senior Staff Engineer at Google Brain. ML systems at planetary scale
  • Built agentic harnesses, sandboxes, and custom tool-call infrastructure years before MCPs existed
VV
Vijay Vasudevan
Co-Founder
  • PhD, Carnegie Mellon University
  • Co-created TensorFlow at Google Brain, the framework running under most of the world's AI today
  • Built AutoML and neural architecture search infrastructure at Google scale
  • Contributed to Waymo's perception and planning systems
  • 10+ years shipping production AI infrastructure from research to real-world deployment

The people who built the AI layer most software runs on, betting on the next one.

Institutional
Coatue Ventures Maven Ventures Hustle Fund WVV Capital
Angel Investors
Andrej Karpathy Jeff Dean Scott Belsky Tom Willerer Brian Shultz
Karpathy wrote the code that inspired GPT-2, and Jeff Dean co-invented MapReduce and led Google Brain. When they back the same infrastructure company, it's worth paying attention to.
Founding Engineer
Backend / Infrastructure
📍 Palo Alto, CA Full-Time Founding Team
Competitive + Equity
Details shared in first call

The control plane doesn't exist yet. The telemetry pipeline, the identity layer, the governance engine: none of it is built. You're joining to make the foundational decisions that will still be running when this is a very large company, working directly with Jiquan and Vijay across product, customers, and the internal agent stack that runs MintMCP itself.

What You'll Do
  • Own the architecture of MintMCP's core infrastructure, identity, permissions, telemetry, governance
  • Build the MCP control and deployment plane that works across all AI systems simultaneously
  • Design and ship the security layer for enterprise-grade agent governance
  • Work directly with customers, understand what they need and go build it
  • Contribute to the internal agent stack, 25+ agents running the business, and growing
  • Define engineering culture and patterns from day one
What We're Looking For
  • 5+ years in backend or infrastructure engineering
  • Deep fluency in distributed systems, you've owned things at scale, not just used them
  • Experience with data pipelines, telemetry, or security infrastructure
  • Track record of shipping production systems that others depend on
  • Genuine excitement about AI agents, you want to build them, not just read about them
  • Startup DNA: strong opinions, bias to action, no need for a complete spec
You're the right person if: distributed systems feel like home, you've owned something that broke at 3am and you fixed it, and the idea of agentifying a company from the inside, while building the platform that lets every other company do the same, makes you want to start on Monday.

Founding GTM
Commercial Lead · First Revenue Hire
📍 Palo Alto, CA Full-Time First Commercial Hire
Competitive + Equity
Details shared in first call

MintMCP is generating $$$ in new revenue week over week, almost entirely inbound, without a single commercial hire. The demand is there, the product is working, and nobody has built the commercial motion yet, so you're the person who figures out how to scale it. You'll talk to CTOs and platform engineering leads about a problem they feel viscerally, which means you need the technical fluency to earn that conversation.

What You'll Do
  • Build MintMCP's commercial motion from zero. ICP, outbound, enterprise sales, PLG
  • Own the full funnel: first conversation to signed contract
  • Define how MintMCP goes to market, strategy and execution, not just one or the other
  • Talk to CTOs and platform engineering leads, earn the technical conversation
  • Work directly with founders; your deals have first-class support from day one
  • Build the foundation for a GTM team as the company scales
What We're Looking For
  • 5+ years selling developer tools, infrastructure, or technical B2B SaaS
  • Experience selling to engineering buyers, you can read an API doc and hold your own
  • Track record of building pipeline from scratch, not inheriting it
  • Early-stage company experience, you know what "figure it out" looks like in practice
  • Genuine technical curiosity, you want to understand the product deeply, not just pitch it
  • Energized by greenfield: an empty CRM is an opportunity, not a red flag
You're the right person if: you've sold infrastructure or dev tools to engineering buyers, you can hold a technical conversation without a SE in the room, and "first commercial hire at a company already doing $$$ in weekly inbound with no sales team" sounds like the best timing you've ever seen.

Small team, hard problem, precise timing.

🎯
The window is open, not forever.
MCP just became the de-facto standard, and the governance layer doesn't exist yet. The first credible team to build it well will own the category; this is a window measured in months, not years.
🤖
Agent-first from day one.
25+ agents already run the business internally, so you won't just build agent infrastructure, you'll use it every day.
📈
$$$ in weekly inbound. Nobody's even selling yet.
Customers are finding us, the product is working, and the commercial motion is still wide open for you to shape.
🏗
Work directly with the founders.
No management layers: Jiquan and Vijay are deeply technical and hands-on, so decisions are fast and your work shows up in production quickly.
🏆
Pedigree that holds up.
TensorFlow, Coursera from idea to millions of learners, Google Brain at scale, backed by the people who built the AI layer running under most of the world's software.
💰
Real equity and upside.
Founding-level equity at a company with real revenue, real customers, and a clear path to becoming critical infrastructure for every enterprise deploying AI agents, tied to a problem every company has right now.