# Hanah

> Hanah is an AI scribe, patient engagement, and practice analytics platform for modern health practices — with regional data residency in Australia, New Zealand and the UK.

## Pages

- [Hanah for allied health — the AI scribe that stays](https://hanah.health/allied-health/index.md): Hanah is the AI scribe that stays after the note — home programs, outcome measures and patient engagement for physio, allied and rehab practices, with regional data residency in Australia, New Zealand and the UK.
- [Benchmark history — Hanah Validation (AU)](https://hanah.health/au/trust/benchmarks/index.md): Every accuracy benchmark run Hanah has published, with the deltas between them.
- [What changed — Hanah Validation (AU)](https://hanah.health/au/trust/comparison/index.md): How Hanah's accuracy benchmark moved between our last published run and the current one, and exactly what changed to cause it.
- [Trust Center — Hanah (AU)](https://hanah.health/au/trust/index.md): Hanah Trust Center (Australia): security and privacy controls, subprocessors, and live model disclosure for the Australian production deployment.
- [Service Status — Hanah (AU)](https://hanah.health/au/trust/status/index.md): Service status for au.hanah.health: per-capability health, 90-day uptime, and incident history for the Australian production deployment.
- [Transparency — Hanah Validation Report (AU)](https://hanah.health/au/trust/transparency/index.md): Hanah's accuracy validation report: per-recording transcripts, generated documents, and per-LLM jury verdicts against an authoritative ground-truth corpus. Full data, not just headline numbers.
- [Validation methodology — WER + LLM jury (AU) — Hanah](https://hanah.health/au/trust/wer-methodology/index.md): How Hanah measures transcription accuracy and documentation hallucination: one harness, one corpus, two complementary evaluation layers — Word Error Rate against verbatim ground truth, and a panel of LLM evaluators against pre-registered clinical fact lists.
- [AI health tech should be transparent. Most of it isn't. — Hanah](https://hanah.health/blog/ai-health-tech-transparency/index.md): Hanah does not pass our own benchmarks with perfect accuracy — across 106 tests it made 3 mistakes, and we're publishing that on purpose. Why we refuse to massage the numbers, and what accountability actually means in health tech.
- [Woven in, not bolted on: our philosophy of AI in healthcare — Hanah](https://hanah.health/blog/ai-woven-in/index.md): Most health software is adding AI the same way: a chat bubble in the corner. We built Hanah the other way around — one agentic assistant, threaded through every surface of the product, with a clinician signing everything.
- [Updating to new model and transcription algorithm. — Hanah](https://hanah.health/blog/benchmark-update-july-2026/index.md): We've updated the model that drafts your notes and the algorithm that transcribes your consults. Our benchmarking shows word error rate down from 5.9% to 1.6%, and invented details down from 3 confirmed cases to none. Here's what changed, and what didn't.
- [New: bring your own template for document generation — Hanah](https://hanah.health/blog/byo-template-docgen/index.md): Upload your existing PDF or Word forms and Hanah fills them straight from the consult — insurance forms, referrals, your own letterhead, ready to review.
- [Setting up your Cliniko API key — Hanah](https://hanah.health/blog/cliniko-api-key-setup/index.md): Two minutes in Cliniko, two minutes in Hanah. Where the API key lives, what Hanah does with it, and how to add the browser integration so you can watch Hanah work from inside Cliniko itself.
- [How to connect Hanah to Cliniko — Hanah](https://hanah.health/blog/cliniko-integration/index.md): Connect your Cliniko account to Hanah in a couple of minutes - your patients and appointments flow in, and your notes go back to the right chart. Watch the full walkthrough.
- [Deploying AI in your practice: six questions to ask before you sign — Hanah](https://hanah.health/blog/deploying-ai-in-your-practice/index.md): Every AI scribe demo looks the same. The differences turn up later — in the invoice, the subprocessor list, and the first consult you record in a room with no signal. Six questions to put to every vendor, including us, with our own answers written down.
- [Hanah vs Heidi: AI scribe vs Allied health oriented workflows — Hanah](https://hanah.health/blog/hanah-vs-heidi-allied-health/index.md): Heidi is an AI scribe. Hanah serves the same practices but keeps working after the note — with the clinical tooling built in. A checkable, source-linked comparison.
- [Hanah vs Heidi: a plain comparison — Hanah](https://hanah.health/blog/hanah-vs-heidi/index.md): What each product does, what it costs, and how each company handles accuracy — side by side, with sources linked so you can check every claim.
- [Hanah vs Lyrebird: AI scribe vs allied health oriented workflows — Hanah](https://hanah.health/blog/hanah-vs-lyrebird-allied-health/index.md): Lyrebird is an AI scribe built around general practice. Hanah is a workflow platform built around allied health. A checkable, source-linked comparison.
- [Hanah vs Lyrebird: a plain comparison — Hanah](https://hanah.health/blog/hanah-vs-lyrebird/index.md): What each product does, what it costs, and how each company handles accuracy — side by side, with sources linked so you can check every claim.
- [Blog — Hanah](https://hanah.health/blog/index.md): Articles and product updates from Hanah — the AI scribe, patient engagement, and practice analytics platform for modern health practices.
- [Educated patients are engaged patients — Hanah](https://hanah.health/blog/inline-health-literacy/index.md): An accurate note the patient can't read is only half a note. Inline health literacy turns the jargon in every visit summary into tappable, plain-language explanations — without watering the clinical note down — because a patient who understands their care is a patient who stays in it.
- [Why the scribe should stay after the note — Hanah](https://hanah.health/blog/scribe-that-stays/index.md): Most AI scribes stop at the note. We think the higher-value work is everything that happens after the consult ends — and that the scribe should stay for it.
- [Hanah for Cliniko — the AI scribe built for Cliniko](https://hanah.health/cliniko/index.md): Hanah is an AI scribe that works inside Cliniko: record on the patient page you already have open, draft into your own Cliniko note template, sync your schedule with an API key, and file finished notes back to the right chart.
- [Data Processing Agreement — Hanah](https://hanah.health/dpa/index.md): Hanah's Data Processing Agreement: the Article 28 processor terms under which Hanah processes clinical data on behalf of clinics in Australia, the United Kingdom and New Zealand.
- [Hanah for hospitals — AI scribe for ward rounds, handovers & discharge summaries](https://hanah.health/hospitals/index.md): Hanah captures ward rounds and consults, then drafts the notes, handovers, discharge summaries and outpatient letters in your hospital's own templates — with regional data residency in Australia, New Zealand and the UK.
- [Hanah — AI scribe, patient engagement & practice analytics for modern health practices](https://hanah.health/index.md): Hanah is an AI scribe, patient engagement, and practice analytics platform for modern health practices — with regional data residency in Australia, New Zealand and the UK.
- [Trust Center — Hanah (NZ)](https://hanah.health/nz/trust/index.md): Hanah Trust Center (New Zealand): security and privacy controls with Privacy Act 2020 and HIPC 2020 alignment, subprocessors, and live model disclosure for healthcare customers and partners.
- [Service Status — Hanah (NZ)](https://hanah.health/nz/trust/status/index.md): Service status for nz.hanah.health: per-capability health, 90-day uptime, and incident history for the New Zealand production deployment.
- [Transparency — Hanah Validation Report (NZ)](https://hanah.health/nz/trust/transparency/index.md): Hanah's accuracy validation report: per-recording transcripts, generated documents, and per-LLM jury verdicts against an authoritative ground-truth corpus. Full data, not just headline numbers.
- [Validation methodology — WER + LLM jury (NZ) — Hanah](https://hanah.health/nz/trust/wer-methodology/index.md): How Hanah measures transcription accuracy and documentation hallucination: one harness, one corpus, two complementary evaluation layers — Word Error Rate against verbatim ground truth, and a panel of LLM evaluators against pre-registered clinical fact lists.
- [Hanah — AI scribe, patient engagement & practice analytics for modern health practices](https://hanah.health/physios/index.md): Hanah is an AI scribe, patient engagement, and practice analytics platform for modern health practices — with regional data residency in Australia, New Zealand and the UK.
- [Pricing — Hanah](https://hanah.health/pricing/index.md): See what your practice gets back. Estimate the hours Hanah reclaims for your team — then pick the plan that fits.
- [Terms, Privacy & Content Policy — Hanah](https://hanah.health/terms/index.md): Hanah's Terms, Privacy & Content Policy: how we handle clinical data, AI human-in-the-loop responsibility, data residency, retention, and your rights in Australia, the UK and New Zealand.
- [AI accuracy & hallucination report — Hanah](https://hanah.health/transparency/index.md): Hanah's AI accuracy and hallucination report: headline Word Error Rate, jury coverage, and confirmed-hallucination counts for every region, rendered from the same published data as each region's full validation report.
- [Trust Center — Hanah](https://hanah.health/trust/index.md): Hanah's Trust Centers: security and privacy controls, subprocessors, and data residency for each regional production deployment — Australia, the United Kingdom and New Zealand.
- [Trust Center — Hanah (UK)](https://hanah.health/uk/trust/index.md): Hanah Trust Center (United Kingdom): security and privacy controls, subprocessors, and live model disclosure for the UK production deployment.
- [Service Status — Hanah (UK)](https://hanah.health/uk/trust/status/index.md): Service status for uk.hanah.health: per-capability health, 90-day uptime, and incident history for the United Kingdom production deployment.
- [Transparency — Hanah Validation Report (UK)](https://hanah.health/uk/trust/transparency/index.md): Hanah's accuracy validation report: per-recording transcripts, generated documents, and per-LLM jury verdicts against an authoritative ground-truth corpus. Full data, not just headline numbers.
- [Validation methodology — WER + LLM jury (UK) — Hanah](https://hanah.health/uk/trust/wer-methodology/index.md): How Hanah measures transcription accuracy and documentation hallucination: one harness, one corpus, two complementary evaluation layers — Word Error Rate against verbatim ground truth, and a panel of LLM evaluators against pre-registered clinical fact lists.
- [Hanah — The AI scribe that stays](https://hanah.health/v21landing/index.md): Stay present during the consultation. Review your notes, referral drafts and patient plans in Hanah, with care that continues between visits.
- [Hanah — The AI scribe that stays](https://hanah.health/v2landing/index.md): Stay present during the consultation. Review your notes, referral drafts and patient plans in Hanah, with care that continues between visits.