Home/Connect/Decagon × Zendesk
The agent
The system

Make Decagon act in Zendesk — safely

Decagon resolves the conversation; Upware carries the action into Zendesk — ticket updated, attributed, and audited before the customer hits refresh.

Request a demoSee the platform
Governed by policy & RBAC Full audit trail Outbound-only on-prem bridge No rip-and-replace
The gap

Decagon stops at the Zendesk boundary

Decagon resolves conversations, but the action that closes the loop — refund, update, fulfillment — lives in systems of record it must reach safely.

Three ways to close it

Only one is governed

Decagon alone
Acts in ZendeskNo
Governedn/a
Auditedn/a
Custom API glue
Acts in ZendeskSort of
GovernedYou build it
AuditedYou build it
Decagon + Upware
Acts in ZendeskYes
GovernedPolicy + RBAC
AuditedEvery action
In practice

From resolution to record, in seconds.

A customer contacts support about a delayed shipment. Decagon confirms eligibility and commits to a resolution — but marking the ticket solved, attaching the internal note, and setting the SLA tag all need to happen in Zendesk with a clear ownership trail. Upware's on-prem bridge replays the exact sequence it learned from a support agent's recorded session: it posts the internal note under the right agent identity, updates ticket status, applies the routing tag, and logs every field change with a timestamp and policy reference. No API had to be written. The compliance team sees a complete audit trail; the customer sees a closed ticket.

The harness

AI to learn. Deterministic to execute.

Upware learns the process once and encodes it as a mostly-deterministic workflow: system interactions replay exactly, and any LLM steps are wrapped in policy, verification, and audit — so execution stays governed and drift-proof.

O
Observe & Learn
Records the real workflow once, across every system it touches.
G
Generate
Builds the API, data flow, and logic automatically — no integration project.
E
Encapsulate
Wraps it as a governed AI skill over secure MCP or API.
R
Run
Executes deterministically — RBAC, audit, and scale built in.
FAQ

Common questions

Decagon already integrates with Zendesk natively — why route through Upware?
Native connectors handle reads and simple writes well, but they don't give you governed execution: no RBAC on which agents can close tickets versus escalate them, no per-action audit trail tied to policy, and no way to reach the Zendesk fields or macros that fall outside the published API. Upware learned those workflows directly from a recorded desktop session and replays them deterministically, so every ticket action is attributable, policy-checked, and logged — something a standard OAuth integration can't provide.
Does Upware put LLM calls inside our Zendesk ticket actions?
The system interactions — updating a ticket, posting a note, setting a tag — are replayed deterministically from a learned recording, not generated by an LLM at runtime. If Decagon uses an LLM to decide what resolution to commit to, that decision happens in Decagon before the action reaches Upware. Upware's job is governed, auditable execution of the resulting action: the right change, in the right field, under the right identity, every time.
How does attribution work when Decagon is the one acting, not a human agent?
Upware's RBAC layer maps AI-initiated actions to a defined service identity with its own permission scope. Every write to Zendesk — status change, internal note, priority update — is recorded in Upware's audit trail with the acting identity, the policy that authorized it, and a timestamp. For SLA reviews or compliance spot-checks, your team can show exactly which action happened, when, and why it was permitted, without digging through API logs or relying on Zendesk's native activity feed alone.

Connect Decagon to Zendesk

See a governed action go from recorded workflow to live in days.

Request a demo