The ITSM platform built on Dexon BPM.
Dexon ServiceDesk is the ITSM platform built on Dexon BPM, PinkVERIFY-certified in 12 ITSM processes — without interruption for 14 years: portal, catalogue, incidents, requests, changes, problems, assets, CMDB, knowledge and service levels. And because Dexon BPM is an executable business architecture, every ticket is traceable to the service, the capability and the indicator it serves: the service catalogue is the architecture’s catalogue, and the service level is an indicator anchored to an objective.
What it is
An ITSM product that includes its platform
Dexon ServiceDesk is built on Dexon BPM, and includes it. Everything a service team needs works from day one — the portal, the ITIL processes, the SLAs, the surveys — and each of them is a governed process on the same platform that runs tax audits and citizen services outside IT. The service a user requests in the portal is the same object Dexon Architect models in the catalogue of services and offerings, linked to the capability and the process that deliver it.
The consequence buyers feel later: when your ITSM needs to grow — a process adjusted to how your organization actually works, a new process that is not IT at all, an integration nobody anticipated — you do not migrate platforms. You extend the platform you already own. That is the difference between buying a service desk and buying a service desk that includes its platform.
The product
What Dexon ServiceDesk ships
The twelve certified processes, working from day one — not as a checklist.
Service operation
Incident, request, problem, change →
The core processes with categorisation, prioritisation, approvals with authority rules, and escalation routes that act before the breach.
Portal and catalogue →
A self-service portal with commercial-language offerings, individual requests and a cart; what users see is the catalogue, and behind every offering runs a process.
Service levels →
SLAs computed against real work calendars, by category, priority and place — each site's calendar and each group's break schedule discounted from the clock — and simulated on a given case before activation, with the trace of every rule evaluated. Priority comes from an impact × urgency matrix, and escalation acts on the percentage of SLA time consumed, with steps before the deadline and after it.
A change calendar that reads the CMDB →
The change board sees every change as a bar in its window on a Gantt, and the platform flags in red the changes that collide on the same configuration item or on one related to it in the CMDB — with whom and on which CI.
Surveys that act →
A low score automatically opens a case, a child of the surveyed one, with the answers as its first comment — dissatisfaction becomes a traceable case, not a statistic. Questions support AND/OR conditional logic, NPS and stars.
Problems that open themselves →
A problem is created automatically when incidents of a category — and, if you want, of a site — reach a count within a time window, and the ones that keep arriving chain to it. Known errors are recorded and linked to the cases they explain; root causes are recorded with their method — Ishikawa, five whys, Kepner-Tregoe — and a major incident carries its dependants with it: its status changes and its comments reach every linked case, and closing it — or a problem — closes them in cascade.
Assets and knowledge
Assets and CMDB →
Discovery, reconciliation, lifecycle and a CMDB where the CI a change touches is the same object the incident points at. Maps are edited in the web with your own relation types and class icons, and every CI opens its own reports.
Knowledge →
A knowledge base with governed articles and deflection at the portal, so what was already solved answers before a ticket exists. Each category's step-by-step procedures — ordered steps with estimated time, links and documents — are maintained in the web, and the technician has them inside the case.
AI, assessed — not just claimed →
Classification trained on your history, deadline-risk prediction, sentiment of every message and grounded answers. AI Capability is one of the twelve certified processes.
Channels
Portal and email →
The self-service portal with its assistant, and email turned into cases with the form, record type and company each mailbox determines. Replies thread to their case by the message headers, the case number or the subject; one email sent to several mailboxes creates one case; auto-replies never open cases; the original message is kept as evidence; Microsoft 365 mailboxes connect over Microsoft Graph with OAuth; and junk mail arrives flagged.
Live agent and Microsoft Teams →
In Teams, users sign in with their corporate identity, get answers with cited sources, confirm the classification before the case is created and receive their case notifications in the same chat (available from November 2026). And a technician can take the conversation from the assistant with its full context, and hand it back when done (available from October 2026).
WhatsApp →
Service over WhatsApp with the same assistant and the same ticket, available from November 2026.
Evidence
The only BPM on the PinkVERIFY list →
Pink Elephant's list is public. Twelve processes, renewed without interruption for 14 years, valid until August 2027 — check it yourself.
Measured on live records →
Backlog, cycle, compliance and satisfaction computed over the operational records themselves, with drill-down to the case. Reports for the service desk, assets, processes and security ship ready by area — including the tickets the predictor flags at risk of breaching their SLA.
Closure quality, audited by sample →
Every closure leaves its facts on the case — who closed it and how, whether the solution was documented, age, SLA compliance, reopenings, and whether the requester had to file again. A monthly sample, the same all month long, repeats the audit on the same cases; reopening a closed ticket requires a reason and leaves its history.
Auditable by design →
Every ticket carries its unalterable event log: who acted, what they decided, on what data, under which rule. And its timeline shows the case history as a Gantt — the SLA window, a bar per status and where the time went — so a complaint is answered without reading the log line by line.
Licences that free themselves →
Consumption of named and concurrent management licences against what is licensed, with automatic release of the licence of technicians who stop signing in — unless they still own open cases or active indicators.
On the roadmap
What arrives next, and when
Each item is available for a demonstration or a proof of concept from the month shown.
Escalation to a live agent
From the assistant to a human agent without losing the conversation.
Microsoft Teams bot
Create and follow tickets and query the knowledge base from Microsoft Teams.
WhatsApp channel
Service over WhatsApp with the same assistant and the same ticket.
Full remote administration from the web
The complete set of remote agent actions over computers, from the web console.
Monitoring alarms configured in the web
Monitoring alarms that create or update tickets, configured from the web.
See a certified process run
Pick a process — incidents, changes, requests. We will show it running end to end: portal, assignment, SLA clock, evidence, survey and the indicator that verifies it.