Email infrastructure built for AI agents.

Create programmable inboxes, receive messages, send replies, process attachments and connect autonomous agents through APIs, webhooks or MCP.

Create Your First Inbox View Documentation

from mailmind import MailMind

client = MailMind(api_key="mm_live_demo12345")

# 1. Give your AI agent a real email inbox in 2 minutes
inbox = client.inboxes.create(
    username="sales-agent",
    display_name="Sales Dispatch Agent"
)

print(inbox.address)
# output: sales-agent@inbox.mailmind.live

# 2. Receive messages & send intelligent replies
messages = client.messages.list(inbox_id=inbox.id)
client.messages.send(
    inbox_id=inbox.id,
    to=["lead@example.com"],
    subject="Re: Product inquiry",
    text="Thanks for reaching out! Here is your custom quote."
)

Built for Autonomous Agent Workflows

Your agent already knows how to think. MailMind gives it a way to communicate.

Instant Agent Inboxes

Generate programmable email inboxes on demand (`dispatch-7f3a@inbox.mailmind.live`) in under two minutes with custom domain support.

Hosted DGX AI Workers

Run classification, intent extraction, summaries, and draft generation on local NVIDIA DGX Spark worker nodes without exposing open inbound ports.

Autonomy & Human Approvals

Four strict permission policy levels from Read-Only to Controlled Automation with human-in-the-loop draft review queues.

Featured Use Case

HouseCall Service Dispatch Agent

See how MailMind powers HouseCall (`dispatch@housecall.site`) — automatically identifying HVAC emergency service requests, checking ZIP code urgency, creating draft responses, and routing technician bookings.

Test HouseCall Demo