Case study — Interactive demo
Service request operations.
A clickable operations console built on local sample data: no auth, no database.
Click through a facility request from submission to dashboard. Sample scenario: HVAC not cooling in Warehouse 2, Dock 4, submitted by the night shift lead. The missing equipment ID routes to the exception queue.
Try the demo
System state
Click any step to jump.
A night shift lead submits a facility request through the public portal. No login required.
Submitted by: Night shift lead, Dock 4
“HVAC not cooling in Warehouse 2. Temperature rising, product at risk. Dock 4 area worst affected. Started around 11 PM.”
Location: Warehouse 2 · Dock 4
Structured fields are extracted from the free text. Each carries a confidence score.
| Field | Value | Conf. |
|---|---|---|
| issue_type | HVAC failure | 0.96 |
| location | Warehouse 2, Dock 4 | 0.94 |
| urgency_signal | product at risk | 0.88 |
| time_reported | ~11:00 PM | 0.72 |
| equipment_id | (missing) | 0.00 |
The request is routed to the correct operational category, not a generic “other” bucket.
category = HVAC AND location CONTAINS "Warehouse" → Facilities / HVAC queuePriority comes from urgency signals, category and time of day.
→ “product at risk” urgency signal
→ HVAC failure in a temperature-controlled warehouse
→ Night submission: on-call path applies
Required fields are checked before a work record is created.
Warehouse 2 has 3 HVAC units and the submitter did not say which. Auto-creation blocked until a person picks the asset.
W2-HVAC-01 Rooftop, north end · W2-HVAC-02 Rooftop, dock side · W2-HVAC-03 Split system, office wingA coordinator selects W2-HVAC-02 (dock side) and the record is created with an audit trail.
SR-2026-0341HVAC not cooling: Warehouse 2, Dock 4
Asset W2-HVAC-02 · Priority High · HVAC / Facilities · Open
The right person is notified based on category, priority and on-call schedule, not a blast to everyone.
○ Facilities supervisor: in-app CC on high priority
The submitter gets the record ID and expected response window.
SR-2026-0341 has been received. HVAC technician R. Chen has been notified. Expected response: within 2 hours.Requests with complete data auto-process; this one waited for a person. The queue shows everything that needs a human decision.
SR-2026-0341 · resolved by coordinator asset selection in step 6
Open requests, exceptions, assignment status and SLA in one view (sample data).
SR-2026-0341 HVAC, Warehouse 2 · HighSR-2026-0340 Lighting, Dock 3 · NormalSR-2026-0339 Door sensor, Dock 4 · NormalProblem
Facility and operations teams get service requests by phone, email, text and walk-up, each with different information quality. Urgent and routine requests look the same until someone reads them. Missing details (which unit, which dock, which asset) only surface after a technician is dispatched. There is no audit trail, no exception path and no acknowledgment to the person who reported it.
Before
A night shift lead texts “HVAC not working in warehouse 2.” The message sits until morning. A coordinator guesses which unit and assigns the wrong technician. Three people report the same issue through different channels and three work orders get created.
The ten-stage workflow
- Request submitted (public portal, no login)
- Information extracted with per-field confidence
- Categorized by routing rules
- Priority assessed from urgency signals
- Missing information identified, which blocks auto-creation
- Record created after a person resolves the gap
- Right person notified (on-call path for high priority)
- Acknowledgment sent to the submitter
- Exception queue for items needing a human decision
- Operations dashboard with open count, exceptions and SLA
It mirrors the public request portal in MaintenEase. This demo shows the deliverable pattern; it is not a production deployment.
What I learned
Block auto-creation on missing required fields. Guessing the asset is worse than waiting.
Acknowledge immediately, resolve separately. The submitter needs to know someone received it, even while the exception is open.
Operations consoles beat chatbot UIs. Coordinators need state panels, exception queues and audit trails, not a conversation thread.
Have a messy process like this?
Or email info@decent4.com · call (618) 606-0066
More case studies