After-hours capture
AI greets the visitor, captures the issue, address, photo, and urgency — even when the team is asleep or on a job. No silence, no missed lead.
Solution · Plumbers and trades
Never miss an emergency callout while your team is in the field
OrcaLinq answers WhatsApp and website messages in seconds, captures address, issue, and photo, and pages the right field tech the moment they're available.
The problem
An emergency callout doesn't wait for a callback.
When a pipe is leaking at 11 PM, the customer messages three plumbers. The first one to respond books the job. Field teams in the middle of work cannot reply fast enough. AI-first answering closes that gap honestly.
Location routing
Inbound is routed to the field team that covers that suburb or service zone. The wrong tech doesn't get paged for the wrong area.
Emergency escalation
Keywords like 'leak', 'flooding', 'no heat in winter', 'gas smell', 'sparking' escalate immediately to the on-call endpoint with a priority flag.
WhatsApp-first
Most trades customers prefer WhatsApp. AI answers and qualifies; the field tech replies from their own phone when they're free.
Conversation example
What a real callout looks like
11:43 PM, the team is at home. Without AI, this customer calls the next plumber on Google. With AI, the lead is captured, qualified, and on the right tech's phone when they wake up.
Customer (WhatsApp, 11:43 PM): "Hi, my kitchen tap is leaking really badly, water everywhere. Can someone come tomorrow?"
AI: "Sorry to hear that. Can you send a photo of the leak and your address? I'll get the next available plumber for your area to confirm a time. In the meantime, please turn off the water at the mains if you can — it's usually under the kitchen sink or near the meter."
Customer: [sends photo + address: "27/B HSR Layout 4th Sector"]
AI: "Got the photo and address. That looks like a fitting failure — definitely needs attention. I'm flagging this as urgent and paging our on-call plumber for HSR Layout. You'll hear back within 30 minutes. The mains shut-off should hold things until they arrive."
Field tech Suresh (replies from his WhatsApp, 6:15 AM): "Hi, Suresh from the team. I'll be at your place by 8 AM. Confirming address — 27/B HSR 4th Sector, near the corner shop?"
Customer: "Yes that's the one. Thank you."
Wake logs: time-to-capture (40 sec), time-to-page (3 min), time-to-tech-confirmation (6 hr 32 min, overnight). The lead was captured at 11:43 PM and warm-handed to Suresh's first morning ping. Without AI: missed call, customer with the next plumber on Justdial.
Recommended setup
What a typical trades business configures
- WhatsApp Cloud API on the team's primary number — the channel customers actually use.
- Website widget on the contact page and emergency-callout page, with the service slug pre-filled.
- Routing policy: postcode/suburb mapping, service-type mapping, emergency-keyword escalation, after-hours rotation.
- Approved knowledge: service areas, basic pricing structure (call-out fee), emergency contact, what to do before the tech arrives (turn off mains, etc).
- Photo intake wired to the tech's WhatsApp so they see the problem before they leave the van.
- On-call rotation in the routing policy so emergency pages go to whoever is on duty, not the whole team.
Honest limits
What this won't do
- It won't quote a price. Job pricing depends on inspection. AI captures, never commits to a number.
- It won't dispatch alone. A human confirms the appointment. AI qualifies the lead, the tech accepts the job.
- It won't fake availability. If no one is on call, the AI says so and offers a callback slot.
- It won't promise SLAs. 'Within 30 minutes' is what the AI commits to for paging — not for arrival. Arrival is a human commitment.
- It won't handle billing disputes. Those force handoff to the owner, regardless of confidence.
Features used
What's doing the work
Sonar
Reads urgency and detects emergency keywords on every inbound message. Floods and gas leaks bypass the queue.
Matriarch
Routes by suburb, service type, on-call rotation, after-hours flag, and customer history. Out-of-area leads escalate to the team lead.
AI human handoff
Forced handoff on emergencies, billing disputes, complaint language, and any conversation outside captured-and-paged.
Knowledge base AI
Service areas, what to do before the tech arrives, after-hours policy, callout structure — all answered from approved knowledge.
Agent-native endpoints
Tech replies from their own WhatsApp on the road. Photo + address arrives on their phone, not a back-office ticket queue.
Wake
Time-to-capture, time-to-page, time-to-arrival logged. SLA performance is measurable across the team.
Metrics that matter
What changes after launch
- After-hours leads captured — typically several per week per van that were going to a competitor.
- Time-to-first-reply — drops from hours to seconds for the AI capture, and from next-morning to within an hour for tech response.
- Photos in hand before arrival — techs arrive with the right tools and parts, fewer return visits.
- Owner phone time — falls dramatically. The owner stops being the dispatcher.
- Lost-lead rate to the next-plumber-on-Justdial — measurable when the customer would otherwise have moved on.
- Job conversion rate from inquiry — improves when the right tech sees the right photo on the right phone.
Trades FAQ
Frequently asked questions
Why does a trades business need this?
Most trades businesses lose money to missed leads after hours and during jobs. Field teams cannot answer the phone with one hand on a wrench. AI captures and qualifies the request; the right person takes over once they're back at the van.
Does it support photos?
Yes — on channels that support media (WhatsApp, Telegram, Instagram, Messenger, the website widget). Customers can send a photo of the leak, the boiler, the broken pipe. The image is stored in tenant-scoped storage and surfaced to the field tech on their routing endpoint.
Can it route by area or service type?
Yes. Routing rules can split inbound by suburb, postcode, service category, urgency, or after-hours flag. Each rule maps to the right field team or on-call person. Out-of-territory leads escalate to the team lead instead of dropping into a void.
What about emergencies?
Emergency keywords ('leak', 'flooding', 'no heat in winter', 'gas smell', 'sparking') trigger immediate escalation to the on-call endpoint with a high-priority flag. Wake records the time-to-page so SLA performance is measurable.
Can it dispatch automatically?
No — and that's deliberate. Job pricing and scheduling need a human eye on the photo and address. The AI captures, qualifies, and pages; the field tech confirms the appointment.
What about price quotes?
AI does not quote prices. Trades pricing depends on inspection of the actual job; misquoting on chat would create disputes. AI captures the request and the field tech quotes after seeing the work.
Can the customer follow up later?
Yes. Conversation memory is keyed by phone number / customer identity. Returning customers get the same field tech where possible; prior job history is visible to the tech before they reply.
Related pages
Keep exploring
See your next emergency callout get captured.
Bring a sample customer message your team would normally miss after hours. We'll show how the routing handles it end-to-end — capture, photo intake, suburb routing, and the page to the right tech.