Email was built for people. We build it for devices and AI agents too.

Why we built it

Teams running device fleets (QA labs, kiosk networks, field hardware, test farms) keep hitting the same problem: shared mailboxes with tangled plus-addressing, a spreadsheet of credentials, a catch-all inbox with forty thousand messages and no way to tell which device a code was meant for.

In Mailcycle a device owns its address. Pairing is a handshake, and revocation is per device.

AI agents hit the same wall. An agent that signs up for things, answers mail or reads receipts needs an address, and handing it your own inbox gives it everything. In Mailcycle an agent pairs like a device, gets addresses of its own and keys for those alone, and starts able only to reply.

One private address stays simple. The individual who wants one, the developer wiring up an agent and the operator provisioning a thousand devices get the same account, the same encryption and the same API. Only the plan differs.

What we believe

• The best way to protect data is not to have it
• Design so you don't have to trust us
• Knowing an identifier is not authorization
• Refuse clearly; never truncate silently
• Mobile and web are the same product
• If the console can do it, the API can do it
• Label planned features as planned
• Say what we can still see
• State what each privacy choice costs