7 daystraining loop live for 20,000 staff, 20 provinces
+50%study-abroad consultant efficiency
80%of repeat questions fielded first
2–4 wksto run one workflow pilot
8industry deployments
2markets: NZ and China
Education & TrainingNZ
From enquiry to enrolment, a funnel that never drops the thread
NZ · International education group · admissions funnel & counsellor collaboration
Enquiry→Application→Offer→Enrolment→Dashboard
✕ BeforeEnquiries scattered across inboxes and counsellors; management saw the funnel once a month.
- Every enquiry auto-files into the funnel; counsellors get follow-up nudges and shared records
- Management opens a live enrolment dashboard any time
Education & TrainingCN
Every counsellor gained a closing co-pilot
CN · Study-abroad consultancy · counsellor assist & knowledge retention
Pre-meeting match→Live prompts→Post-meeting profile→Case library
✕ BeforeMatching meant digging through files; closing know-how lived in senior heads and left with them.
- Auto-matched country and program options pre-meeting; talk tracks and key questions live
- Post-meeting student profiles and follow-ups flow into the firm’s case library
Construction & TradesNZ
Payables that no longer need chasing
NZ · Construction company · payables & payment alerts
Invoice OCR→Approval nudges→Payment tracking→Overdue alerts
✕ BeforeInvoices keyed by hand, approvals stuck with no visibility, overdue only discovered after the fact.
- Agents read fields, chase approvals and payments, and warn before anything goes overdue
- Approvals and project accounts on one finance dashboard
Full case study: 91% invoice automation →
Construction & TradesNZ
Quotes go out only after the details come in
NZ · Roofing company · quoting & project intake
Enquiry intake→Info chase-up→Project card→Warranty reminders
✕ BeforeEnquiries jotted anywhere; quotes missing measurements and photos; days lost to back-and-forth.
- Agents chase a checklist of details, then generate the project card and follow-up scripts
- Site photos archived; warranty expiries trigger automatic reminders
Trade & LogisticsNZ ⇄ CN
WeChat-group orders, closed-loop for the first time
NZ ⇄ CN · FMCG export · order loop & reconciliation
Group order→Payment screenshot→Finance confirms→ERP write-in→Tracking no. returned
✕ BeforeOrders and payment screenshots buried in group chats; support scrolled back message by message; month-end never balanced.
- Agents extract order details and match payment screenshots; finance confirms, ERP updates
- Tracking numbers post back to the group; the owner’s dashboard stays current
Trade & LogisticsNZ ⇄ CN
An order desk that never drops a shipment
NZ ⇄ CN lane · freight forwarder · order progression & exception alerts
Structured enquiry→Quote decision card→Order board→Exception alerts
✕ BeforeEnquiries scattered across WeChat, quotes resting on one person’s judgment, missed follow-ups meaning lost orders.
- Enquiries auto-structured; internal decision cards support quoting
- Order board plus exception alerts watch every shipment; follow-up scripts ready to send
Finance & Business ServicesNZ
Every client milestone, remembered
NZ · Mortgage adviser team · client lifecycle & CRM activation
Voice-to-file→Milestone nudges→Meeting brief→Case tracking
✕ BeforeRefix dates, birthdays, settlements — every client touchpoint depended on an adviser remembering.
- Refix reminders, birthday emails and settlement congratulations fire on lifecycle triggers
- Speak once and the file is created; a one-page brief lands before every meeting
Finance & Business ServicesCN
No CRM swap — an agent layer grew on DingTalk
CN · Venue & park leasing · pipeline management & daily ops brief
Message structuring→Sales confirms→Pipeline table→Owner’s daily brief
✕ BeforeFollow-ups lived in chats and heads; CRM fields sat empty; progress required calling a meeting.
- DingTalk texts, voice notes and screenshots auto-structured, written to the pipeline after confirmation
- The owner gets a daily operations snapshot
What AI implementation case studies do you have in New Zealand, and how are results measured?
Eight delivered projects across New Zealand and China: an admissions funnel for an international education group, payables automation for a construction firm, quote intake for a roofing company, WeChat-group order loops for cross-border logistics, client lifecycle automation for mortgage advisers, and an agent layer on DingTalk for a business-park operator. Each case states the manual baseline and what changed; acceptance metrics such as repeat-question absorption, document turnaround and human-review pass rate are written before work starts.
Why are the cases anonymised, and can we see full versions?
Clients asked us to withhold company names while keeping the full process detail. Tell us your industry when booking a diagnosis and we’ll walk you through the complete case for that sector, including the metrics.
We’re not in one of these eight industries. Is this still relevant?
Yes. The industries differ, but the patterns repeat: data scattered across chats and inboxes, follow-up that depends on memory, reconciliation done by hand. The agents connect to the systems you already use, so the workflow logic transfers; only the vocabulary changes.
How long from a case like this to our own go-live?
Follow-up, reconciliation and knowledge Q&A pilots typically close a minimal loop in 2–4 weeks; complex scenarios expand workflow by workflow over 3–6 months. Every step has acceptance criteria and a way back.