Cunav — Support & Ticketing
Support that triages itself
Cunav handles bug reports, feature requests, and questions the way any support tool does — queues, tickets, comments. The difference is what happens the moment a ticket lands: AI can read it, judge it, and act, before a person even opens it.
Why Cunav
Support, wired into the same engine as everything else
A Cunav queue is an AWE Job. A ticket is an AWE Workflow. That's not an implementation detail — it's why support can plug directly into delivery instead of living in its own silo.
AI Triage on Arrival
Flag a queue as AI-enabled and every new ticket gets read and judged automatically — with a proposed next step and a confidence score, not just a label.
Confirmed Work Routes Itself
Above your confidence threshold, Cunav can create and link a Togra story automatically — a real ticket becomes a real piece of scheduled work, with no one copying it across by hand.
Agents Can File Their Own Tickets
An AI agent — yours, a customer's, or ours — can report an issue or check its status through a dedicated service account, without ever needing a Cunav login.
Familiar Ticket References
Tickets are numbered like any tracker you've used — ULLAV-42 — and those references resolve automatically in commit messages and pull requests over in Lagan.
Threaded Comments, Full History
Every reply, decision, and status change is recorded against the ticket, shared with the same notes infrastructure the rest of the platform uses.
Built on the Workflow Engine
Nothing about Cunav is a bolt-on integration — queues and tickets are first-class AWE Jobs and Workflows, so every capability of the engine is available to support work too.
AI Triage
Let AI read the ticket, before a person has to
This is the headline reason Cunav exists in its current form: incoming support work doesn't have to wait for a human to open, read, and categorise it first.
Judged, not just sorted
Triage isn't keyword routing. Each ticket is read for what it actually says, and Cunav proposes a next step — not a bucket.
Confidence-gated action
Below the confidence threshold, a ticket waits for a human. Above it, Cunav can act on its own proposal — you decide where that line sits, per queue.
Routes straight to delivery
A proposal like "this is real work" can create a linked Togra story directly — support and delivery share a workflow, not a copy-paste habit.
Nothing hidden
Every triage decision — proposal, confidence score, and outcome — is recorded on the ticket, auditable the same way any other workflow step is.
Built for Agents Too
Your AI agents can talk to support directly
Cunav exposes a small set of tools any AI agent can call — report an issue, check on one, look up a ticket by reference — without ever logging into Cunav itself.
Report an issue
An agent working on your behalf can file a ticket the moment it hits a problem, with full context attached — no context-switch back to a human.
Check ticket status
"What's the status of ULLAV-42?" works for an AI agent exactly the way it works for a person.
A dedicated service identity
Agent access runs through its own service account, kept separate from any individual user's login or permissions.
One more workflow surface
This is the same pattern AWE uses everywhere — the workflow engine is as accessible to an AI agent as it is to a person at a keyboard.
How It Works
From ticket to shipped fix
A typical ticket's path through Cunav and the rest of the platform.
A ticket lands
Filed by a customer, a teammate, or an AI agent through the Cunav MCP tools — no login required for the agent path.
AI triages it
On AI-enabled queues, the ticket is read and judged automatically, with a proposed next step and a confidence score.
Confirmed work routes to Togra
Above the confidence threshold, a linked Togra story is created automatically — the ticket becomes scheduled work.
Referenced all the way to the fix
The ticket reference resolves automatically in the commit and pull request that close it over in Lagan — the whole path stays traceable.
Questions?
Frequently asked questions
What does "AI-enabled queue" actually mean?
It's a flag on a queue. When set, every incoming ticket in that queue is automatically triaged — read, judged, and given a proposed next step with a confidence score — before or alongside a human looking at it.
What happens when the AI gets it wrong?
Confidence gating means low-confidence proposals simply wait for a human — nothing acts autonomously below your threshold. Every triage decision is recorded on the ticket, so a wrong call is visible and reviewable, not silent.
Do our own AI agents need a Cunav account to file tickets?
No. Agent access runs through a dedicated service account and a small set of MCP tools — report an issue, check status, look up a ticket — kept separate from individual user logins.
How does a Cunav ticket connect to Togra and Lagan?
A confirmed ticket can create a linked Togra story automatically, and the same ticket reference resolves inside commit messages and pull requests in Lagan — one thread from report to fix.
Can we migrate our existing support queues onto Cunav?
Yes — this is exactly the kind of tool integration our consultancy engagements cover, including deciding which queues are worth AI-enabling first.
How is Cunav priced and deployed?
Cunav is offered on a project and agreement basis, hosted, on-site, or hybrid on Kubernetes. Get in touch to discuss your support volume and we'll put together a proposal.
Ready to let AI take the first pass on support?
Tell us about your ticket volume and let's work out where AI triage saves your team the most time.