Startups & Funding
AgentMail raises $6 million to build email infrastructure for AI agents
AgentMail raised $6 million in seed funding to provide an API platform that gives AI agents email inboxes, positioning email as a core identity layer for autonomous software.
On Tuesday, San Francisco-based startup AgentMail announced it has raised $6 million in a seed funding round. The company provides an application programming interface (API) platform—which allows different software applications to communicate with one another—that gives artificial intelligence agents their own email inboxes. This infrastructure is designed to move beyond the traditional human-centric email experience, allowing autonomous software programs to send, receive, and manage emails programmatically.
The startup has experienced rapid adoption, particularly following the late January debut of OpenClaw, an AI agent platform. While coding agents like Claude Code, Codex, and Cursor initially drove adoption among programmers around the world, OpenClaw’s release further accelerated the demand for AI agents. According to co-founder and CEO Haakam Aujla, AgentMail’s user count quadrupled in February, and the platform has now attracted tens of thousands of human users, hundreds of thousands of “agent users,” and more than 500 B2B customers. Instead of requiring agents to navigate graphical user interfaces designed for humans, AgentMail allows them to interact with software services directly through API calls. “We thought we wanted our agents to be able to do that, but they shouldn’t have to, you know, click buttons on a screen, because that’s pretty clunky for agents to do. They should just be able to make API calls,” Aujla said.
The seed funding round was led by General Catalyst, with participation from:
- Y Combinator (which AgentMail launched with as part of its Summer 2025 batch)
- Phosphor Capital
- Angel investor Paul Graham
- Angel investor Dharmesh Shah (CTO of HubSpot)
- Angel investor Paul Copplestone (CEO of Supabase)
- Angel investor Karim Atiyeh (CTO of Ramp)
Beyond basic communication, Aujla frames the service as an identity layer for AI agents. By assigning an email address to an agent, it can use existing internet infrastructure to register for and interact with almost any software service in the world. However, giving email inboxes to autonomous software raises concerns regarding potential “misuse” and “abuse.” To counteract these risks, AgentMail has implemented safety controls, including a strict limit of 10 emails a day for unauthenticated agent inboxes, alongside rate limits and automated monitoring for unusual activity.
Why it matters
AgentMail is attempting to solve the identity problem for AI agents by leveraging existing email infrastructure, allowing agents to interact with software services without needing human-like user interfaces.