Support that closes the loop—not just suggests the next sentence.
Why we're building this
At our last companies we watched support orgs stitch together ticketing, automation, and a chat widget — and still ship a janky experience to the customer.
Most “AI for support” tools today suggest replies and stop there. The customer still waits. The agent still copy-pastes. We wanted something that closes the loop: looks things up, takes the action, and escalates with full context when it should.
How we work
- Distributed team, building in public on a tight loop.
- Policy and approval gates are first-class — humans stay in the loop on the actions that matter.
- Conversation and ticket history, with published workflow snapshots.
- We ship the dashboard our own ops team would want to use.
What we optimize for
These are the calls we make when there's a tradeoff between “magical demo” and “works in production.”
Action over suggestion
Copilots that suggest text leave the customer waiting. We close the loop end-to-end, with gates when risk is high.
Policy at the edge
Project access, tool permissions and confirmation requirements shape the actions an agent may take.
Earned, not claimed
We distinguish current capabilities from planned work, and label illustrative examples separately from customer results.
Try it, or talk to us.
The fastest way to know if Subport fits is to load your knowledge base and run a few real tickets through it. Reach us at hi@subport.io if you want a hand.