Pilot purgatory
The demo wowed the board. Nine months later it's still only on one site, the vendor is "working on" the integration, and the rollout slide hasn't moved.
We've taken pilots stuck at 1 site to 100+ sites in under 12 weeks.
We're the independent advisory for GP corporates, aged-care operators, dental groups and imaging networks running 10+ sites. We pick the right AI voice platform for your network, run the deployment, and hand you an AI receptionist that books, reschedules and writes back to your PMS — evaluated against our published CAPR framework, not another smart voicemail your front desk has to clean up.
Cadence — independent AI voice advisory for ANZ healthcare.
Engagements have spanned GP corporates, allied-health groups, imaging networks and aged-care operators across ANZ.
We do not sell single-clinic point solutions. Our engagements are network-wide evaluations for operators who need one platform that works across every site.
The vendor evaluation framework we built during that engagement, packaged for healthcare operations leaders. Skip three months of vendor demos. Free white paper, no gate beyond your work email.
8 domains · 24 pages · used to choose the vendor now live across clinic networks we advise
"We tried an AI voice agent. It didn't bring the results we hoped for. What did we do wrong?" — a question we get on almost every fit call this year.
The demo wowed the board. Nine months later it's still only on one site, the vendor is "working on" the integration, and the rollout slide hasn't moved.
We've taken pilots stuck at 1 site to 100+ sites in under 12 weeks.
Patients only offered 5 of 15 doctors. Bookings hallucinated. 30% ask for a human the moment they realise it's AI*. Front-desk turned it off by week three.
We audit failed deployments and tell you whether to fix it or replace it.
Invoice is in the system. Renewal is coming up. Nobody measured before, during or after — so no one can answer "how many extra bookings did we get?"
We benchmark every deployment against ANZ peers so the next board meeting is short.
* Metrigy consumer survey, Q1 2026 (n=1,000). Failure patterns sourced from MedCity News, Healthengine, and ForHealth case material, 2026.
Most advisories hedge. We don't. If the Diagnostic doesn't deliver, you don't pay. If you proceed, the fee disappears. If we have a vendor relationship, you read it on the page.
If we can't shortlist a viable platform for your network inside the 2-week Diagnostic, we refund the $7,500 in full. No negotiation.
Proceed to Selection & Rollout within 60 days and the Diagnostic fee is credited 1:1 against the engagement.
We disclose any commercial arrangement we have with any vendor we recommend. If we can't disclose it, we don't recommend them.
Vendor licence and call fees are paid directly by you to the platform we select — no resale markup from us. We run a maximum of 3 concurrent network engagements — capacity is confirmed on the fit call.
Two-week vendor-fit assessment. Board-ready memo in 15 minutes of reading.
We pick the vendor, negotiate the contract, and run the deployment across your sites.
Ongoing oversight for networks live on AI voice. Performance, benchmarks, renewals.
Full inclusions, bonuses, and the Diagnostic refund guarantee are shared in the fit call.
Numbers tuned to ANZ primary care benchmarks — your network's actual figure lands inside the 2-week Diagnostic, alongside the vendor shortlist and a 12-month TCO. Run the math on your own network with the ROI calculator.
No twelve-month implementation cycles. We pilot fast, measure rigorously, and scale what works.
We learn how your network operates today, where calls are being lost, and where AI will deliver the fastest return.
Go live with 5-10 clinics and measure what matters — calls answered, patients booked, staff time saved.
Once it's proven, we roll it out across your full network. Your team gets hands-on training and ongoing support.
Your AI gets smarter every month. We monitor performance, fine-tune, and report back quarterly so you always know the ROI.
Our policy: healthcare only. Every evaluation criterion, integration spec and compliance overlay we publish is purpose-made for clinical environments — not retrofitted from a horizontal SaaS bench.
Every call answered. Every recall sent. After-hours covered. Designed for how Australian general practice actually works — RACGP accreditation, Medicare billing, and the challenge of running 10, 50, or 100+ sites.
Fill more chairs, reduce no-shows, and give your central team back hours they're currently spending on the phone. Built for multi-location dental service organisations.
Families get answers faster. Staff spend less time on the phone. Documentation meets the new Aged Care Act 2024 requirements without adding headcount.
Urgent cases get triaged instantly. Routine bookings happen without staff. Vaccination reminders go out across every clinic in your network.
Patients get booked faster, referrals don't get lost, and prep instructions are delivered automatically — across every location.
One of Australia's largest multi-site GP networks came to us after a failed 6-month engagement with their first platform. We ran an independent evaluation and helped them select the vendor now live across the network.
An enterprise text-chatbot vendor failed after 6 months. 4-second response times. Medical terminology accuracy below 80%. The board needed an independent read on what to do next.
We ran an 8-platform evaluation against Australian data residency, conversation latency, PMS integration risk, and 12-month TCO — then handed the board a single written recommendation naming the vendor that fit.
The board selected our recommended vendor. Pilot clinics live within weeks; rollout now across clinic networks we advise handling production call volumes at the published latency bar. The vendor runs the deployment; we provide ongoing oversight.
Network anonymised. Deployment is run by the selected vendor under Cadence oversight; rollout is ongoing. Detailed reference available under NDA after a fit call.
We're not consultants who've read the vendor brochures. We evaluate these platforms against how ANZ clinics actually run — and we publish the failure modes before we recommend a shortlist.
We evaluate AI voice platforms against how Australian clinics actually run — PMS quirks, AHPRA, after-hours triage, language-barrier callers. That's the only category we work in. Every scorecard weighting, integration spec and compliance overlay is purpose-made for clinical environments, not retrofitted from horizontal SaaS.
Our policy: healthcare only, three network engagements per quarter — we don't take retail, hospitality or legal.
We evaluate every major ANZ-serving platform against a published bar — latency, PMS write-back depth, ANZ data residency, concurrency ceilings and code-switching handling.
Our deliverables cite the specific Privacy Act 1988 clauses, AHPRA guidance, and Aged Care Act 2024 obligations that bind each vendor. Data residency is checked, not assumed.
Diagnostic in 2 weeks. First live site in 6. Multi-site rollout governance after that. No 200-page deck, no quarterly working group, no rebadged Big-4 graduate running your project.
Every publicly documented AI voice, ambient scribe and contact-centre deployment across Australian and New Zealand healthcare — with the source for each. Open JSON + CSV, CC BY 4.0.
318 named ANZ healthcare operators. 15 (4.7%) in a buying window. Machine-readable, CC BY 4.0.
11,600 words. Every figure cited. The independent evaluator's report for multi-site healthcare networks — no paid placements.
Our advisory practice is ANZ-focused. For US or UK networks we take a limited number of remote engagements — start with the 2026 Landscape Report and mention your region.
Start with a 30-minute fit call. If your network isn't ready, or we're not the right team, we'll say so on the call — no deck, no follow-up sequence.