Service desk
One queue for staff, tenants, customers, engineers, and imported demand.
- Find requests and keep context in view
- Service shortlist for the agent to confirm
- Priority and routing on the record
Features
Rules your team can maintain, one auditable work record, and AI that stays under oversight when you use it — not another bolt-on chatbot or consultancy rewrite.
01Operating model
The old CAFM pattern hid business rules inside custom objects, scripts, integrations, and consultancy change requests. AJAEX keeps the rules visible so the platform can move as the operation changes.
01
Sites, spaces, assets, services, permits, documents, contractors, and requests sit on one live record model.
02
Teams configure statuses, SLAs, evidence rules, approvals, access, and escalation paths without burying logic in custom code.
03
Where AI assists, it stays scoped to the record, the allowed capabilities, and rules your team approve — not a free chatbot.
04
Low-confidence or risky outcomes go to people. Every input, output, model, confidence score, route, and correction is retained.

02Platform capabilities
Instead of listing every setting, the platform is easier to understand as eight connected capability groups.
One queue for staff, tenants, customers, engineers, and imported demand.
Start with who or where; structure the request before it enters the queue.
Estate, asset, schedule, job history, and compliance state connected.
High-risk work controlled by permit rules, evidence checks, and review gates.
Documents become structured records instead of files people have to chase.
Business logic stays visible, configurable, and auditable.
Bring legacy data in and push clean operational updates back out.
Assign field work with skills, windows, and day routes in mind.
03Despatch
Field teams get work assigned with skills, time windows, and routes in mind. The film shows the despatch capability we are building: planning a full day of visits instead of chasing the next ticket alone.
04Guided intake
Guided request logging can start with who or where — person, place, then problem — with a structured read-back before the request enters the queue. The film is actual product UI with demonstration data.
05Service desk
Find a request in the queue and keep context beside it, then open the full record with status in view. The film is actual product capture with demonstration data — not a claim of live enterprise load or parity with another platform.
06AI suggestions
Practical help at the service desk — AI can suggest options while human review stays in the flow. In the film, the AI workflow beat is labelled illustrative because that live task was unavailable; product screens use demonstration data.
Governed AI
Users are not left chatting with a model. AJAEX loads the right record, context, rules, and response schema for the task. The AI checks the work, returns structured output, and the workflow decides whether it can proceed or needs human review.
The desk uses site, contract, location and available services to shortlist options. The agent confirms — nothing auto-applies without a person.
Document handling can extract key fields and link evidence to assets or work. Low-confidence or unclear results stay with people for review.
Method statements sit with the permit record so reviewers can check hazards, site rules and controls before work progresses.
AI transaction record
No memory. No broad API access. The record shows what was sent, what came back, and what happened next.
Human oversight stays in the loop.
Low-confidence, incomplete, or risky outcomes can be stopped for review before they affect live records.
07Operation lines
AJAEX covers the common CAFM backbone, then lets each customer switch on the lines that match their operational risk.
Sites, buildings, spaces, assets, and work orders in one work record.
Maintenance schedules, job cards, and asset health.
A full register over a hierarchical estate, site to asset.
One queue for every kind of request.
08Guardrails
Audit, access, residency, and AI transparency are not separate consultancy workstreams.
Every record, comment, document, status change, AI transaction, and correction has a traceable history.
Inputs, model, route, output, confidence, and review outcome are visible on the transaction record.
Starter and Professional tiers run on AWS eu-west-2 (London). Enterprise can pin to a chosen AWS region.
Users only see the records, sites, services, and capabilities available to their role and context.
Bring one messy workflow: a permit, inspection pack, contractor handoff, or service request. We will show how it becomes rules, transactions, review, and evidence.