Built for the whole therapy week.

Anchr for Therapists is the AI-assisted platform built for the whole therapy week - not just the hour your client is in front of you. The portal holds your caseload, schedule, telehealth video, session notes, formulations, screening trajectory, safety plans, MHCP / GP letters, the consent ledger and the billing layer in one place. Anchr Assistant (Cmd-K from anywhere) reads the chart and drafts the admin for you. Anchr Guide and Talk it Through live on your client's phone between sessions, paced through the modules you assigned and feeding every reflection back to the chart.

It's an EHR, an AI scribe, a homework engine, a between-session co-therapist, and a personal admin assistant - fused into one record so context never has to be reconstructed.

Australian psychologists, psychotherapists, mental-health social workers, counsellors and clinical registrars in private practice or small clinics. The platform is opinionated about the AU regulatory context - AHPRA, the Privacy Act 1988, Better Access / MHCP, AU data residency - so it's most at home there. The clinical surface itself (six modalities, MBC engine, safety plans, supervision) is jurisdiction-agnostic; the operational scaffolding (MHCP session counters, GP review letters, CPD registration year) is AU-first.

Solo practitioners through to mid-sized group practices are the sweet spot. Larger multi-clinic admin (org-wide reporting, claims submission, e-prescribing) is on the roadmap rather than in the box.

Anchr is the clinical EHR - chart, notes, formulation, screening, safety planning, MHCP, consent, ROI, AHPRA-compliant chart export - plus the scheduling, telehealth, and between-session surfaces that traditional EHRs treat as out-of-scope. Most of our therapists either replace their old EHR entirely or run Anchr alongside it during a transition.

What we don't do (yet) is health-fund claims submission, e-prescribing, and multi-clinic billing administration (cross-clinic reporting and group billing across many therapists). If you need Medicare bulk-billing claim transmission, Halaxy still does that better than we do today. For everything else - calendar, telehealth, chart, AI scribe, homework, MHCP referrals and review letters, the cancellation-fee ledger, and direct client charging via Stripe Connect - Anchr is built to be the one place the practice runs.

14 days, no card required. The trial is the full platform - telehealth, AI scribe, Anchr Assistant, the 276-module library, the consent ledger, MHCP referrals, the cancellation-fee ledger, direct client charging via Stripe Connect, chart export. Bring real client work into it. The only constraint is seat count; trial accounts default to a single therapist seat, and you can invite supervisors / peers for review during the trial without burning seats.

If you sign on within the trial period, the first 200 founding members get $99 / month locked in for life (the standard plan is $149 / month). See the pricing section below for what's included.

You can self-onboard in about ten minutes - sign up, create your therapist profile (credentials, AHPRA number, bio, default appointment length, default modality), and start sending client invites. The first time you click + Create Event the platform walks you through the difference between a Client Session and a Personal Event.

For practices migrating from another EHR, we run a guided onboarding call (no charge during the trial) where we help you import your caseload, set up your cancellation-fee policy and consent templates, and configure your between-session module defaults. Book the demo from the landing page; we'll send a calendar link.

Four AI surfaces, all grounded in the same chart:

• Anchr Assistant - Cmd-K orb in your portal. Reads the chart, plans sessions, drafts notes / emails / letters, edits your calendar. Every write action queues for one-click confirm.
• Brainstorm - pre-session studio for you. Working formulation, falsifiable techniques with literal opening lines, conspicuous absences in the record. Pin the bits that matter onto a one-page Session Brief.
• Anchr Guide - between-session self-help conversation for your client, AI co-pilot that walks them through the modules you assigned (their words, their data).
• Talk it Through - the same client-facing AI but in conversational chat mode, phase-aware across CBT/DBT/ACT/Psychodynamic/IPT. Every Tuesday morning, the threads that resonated and bounced sit on your chart in the client's own words.

The Assistant and Brainstorm are for you; Anchr Guide and Talk it Through are for your client. All four share the same record - the AI never sees less than you do, and it never sees more than your client consented to.

Three classes of action:

Read tools (instant, never write): caseload triage, compliance scan, "what needs attention", "stuck clients", "ready to graduate", today's schedule, full chart spine for any client, pre-session brief, MBC trajectory, safety status, recent session notes, live telehealth transcript (in-call only).

Write tools (queue as pending action, paused for your confirm): clear an afternoon, create / reschedule / cancel an appointment, create or remove a personal event, draft a session note from a voice memo, draft a client email, draft a summary blurb, queue homework modules, draft an MHCP review / discharge / referral-back letter. Nothing lands in the chart or on the calendar until you tap confirm.

Deep-think jobs (long-context reasoning, async): brainstorm a tough case, run a supervisor critique on a draft note, scan a note for compliance issues. Progress shows in the assistant panel; you keep working.

No. Every write action from Anchr Assistant queues as a pending action - you see exactly what will happen, you tap confirm, only then does anything change. You can reject, edit, or re-issue at any time.

The one exception is extractions from a signed session note: when a recording is processed, the AI labels themes / techniques / quotes against the note and stores them as a separate extraction layer. The note itself is unchanged. Extractions show alongside the note with confidence scores and quote anchors, and you can accept / reject each one. Your accept/reject feeds back into our calibration so future extractions are more accurate for your style.

Anchr Guide is the AI on your client's phone - their self-help companion inside the modules you assigned. It walks them through the lesson at hand, in their own data, in plain language. It reads the modules you assigned, the lessons your client has finished, every reflection they've written, and their mood / defusion / Living Compass markers.

It does not see your session notes, your formulation, your MHCP referral, your private supervisor critique, the consent ledger, or anything you haven't shared explicitly. The wall between "what the therapist holds" and "what the client's AI co-pilot can see" is hard-coded and RLS-enforced at the database layer.

Same context window, different surface. Anchr Guide lives inside a lesson - if your client is doing the Catastrophising Worksheet at 11pm on Tuesday and gets stuck, Guide walks them through the next step using their own data.

Talk it Through is the always-on conversational chat. They can open it from the home screen, and it's phase-aware, integrative across CBT/DBT/ACT/Psychodynamic/IPT, drawing on ~170 evidence-based techniques. Every conversation gates on a Session Card, signed Summary, and evolving Case Formulation. Crisis language is never silently filtered - it surfaces.

By Tuesday morning, what resonated, what bounced, and unresolved threads sit on your chart, in your client's own words.

The Assistant is fast and tactical - "draft this email", "clear my Friday afternoon", "show me Sarah's MBC trajectory". Brainstorm is slower and deeper. It's the pre-session studio that reads the entire client record (mood, defusions, lesson notes, screening trajectory, Talk it Through transcripts, your signed notes), pulls a working formulation, returns falsifiable techniques with literal opening lines, and surfaces the topics that are conspicuously absent from the record.

You pin the bits that matter onto a one-page Plan / Context / Watch For brief. You walk in already five minutes deep instead of catching up for fifteen.

Claude Sonnet 4.6 and Claude Opus 4.x for clinical reasoning; Deepgram Nova 3 for transcription. Both are accessed through enterprise APIs with zero data-retention agreements - your client's transcripts, your notes, and your conversations with Anchr Assistant are not used to train the underlying models.

Every AI call is logged to a tamper-evident ai_call_audit table (model name, prompt template version, input token count, output stop reason, calibration confidence). If you ever need to produce a paper trail for AHPRA / Privacy Commissioner, it's there.

Telehealth runs on Daily.co with the Australian region pinned (geo=au). Rooms are private, knocking is on, and each session gets a unique room URL with a 60-minute access token. Your client clicks a single magic link in their email or push notification - no app install, works on web / iOS / Android in the browser.

You see a waiting-room indicator when they join; tap Admit when you're ready. The session state machine tracks waiting → in-call → ended → finalised, and if anything goes wrong (network drop, recording missing, transcription stuck), the UI shows you exactly which step failed and why - never a silent "Generating note…" forever.

No. Recording is off by default and gated by an explicit two-sided opt-in. You enable recording for an individual client by setting their consent scope; your client also has to tick the recording checkbox on their consent page (the same page where they agree to telehealth in the first place). Until both have happened, the AI scribe and post-session note drafting are unavailable for that client.

Recording is audio-only by default when enabled, because clinical notes don't need video and audio-only halves the data footprint. We capture the recording inside the Daily-managed AU region, transcribe it via Deepgram (also AU-resident), and pipe it through the post-session pipeline. The raw recording is retained for 90 days then auto-purged; the structured note + extraction stays with the chart.

From the recording, four artefacts:

• Transcript - speaker-diarised, timestamped, searchable.
• Draft session note - pre-populated against your chosen template (you can pick from SOAP, DAP, narrative, or your own). You read, edit, and sign.
• Structured extraction - risk markers, homework assigned, client-stated goals, affect observations, values alignment, medication mentions, quote highlights, session themes, techniques used. Each item is quote-anchored back to the transcript moment and confidence-scored.
• Client-facing summary - short reflective summary your client sees in their app post-session, scoped to what's safe to share.

If the AI is unsure about a speaker attribution or a clinical claim, it's labelled unverified rather than asserted. The validator pass demotes confidence rather than deleting items, so nothing the model heard gets silently dropped.

Yes. Tap Record on the appointment row in the portal (or from the iOS app); your phone or laptop mic captures the room audio and uploads it to the same AU-resident pipeline as telehealth. Consent posture is identical - the same recording-consent toggle on your client's consent page gates in-person recording too.

If you don't want to record in person, you can still create a placeholder session note and dictate it after, photograph a handwritten note (Anchr OCRs and structures it), or just type it like any other note. The AI scribe is an option, not a precondition.

Yes. From any client's chart, tell Anchr Assistant "draft a note from this voice memo", drop in the audio, and within 30-90 seconds you'll see a draft note pre-populated against your template. Same applies to a photo of a paper note - OCR + structure into the same signable record.

Transcribe-the-call, dictate-after, and snap-a-paper-note all funnel into the same end state: a structured, searchable, signable note in the chart.

A note is signed once. Once signed, the body becomes immutable for AHPRA / Privacy Act audit purposes - you can't quietly edit yesterday's note tomorrow to change a clinical claim.

An addendum is a timestamped, signed amendment that attaches to the original note. If you remember something that should have been there, or you need to correct an error, you write an addendum - the original stays intact, the addendum is captured separately, and an auditor (you, your supervisor, a regulator) can reconstruct exactly what was known and when.

Yes. Invite your supervisor to the platform; they get scoped, read-only access to specific clients you nominate. They can request changes, leave structured critique comments on notes, and (with your permission) trigger the Supervisor Critique AI tool that runs a deep-think pass over the note and flags formulation gaps, risk-screening omissions, and clinical-reasoning weak spots.

The supervisor never has write access to the chart - only you can sign a note or amend a formulation. Their critique is captured in a separate supervision events ledger that satisfies AHPRA's documentation requirements for registrar / endorsement pathways.

You tap reject on the item. It's gone from the surfaced extraction, the system logs your rejection against the prompt version that produced it, and we use that signal to calibrate future extractions for your style and patient cohort. The underlying transcript is untouched - the AI's interpretation is a separate layer from the source material.

For high-stakes items (risk markers, medication mentions) we surface even low-confidence hits and clearly label them as such - we'd rather you reject a false-positive than miss a real one. Quote-anchoring lets you click any extracted item and jump straight to the transcript moment that produced it.

276 modules · 1,201 lessons across six modalities: CBT, DBT, ACT, Psychodynamic, IPT, and a General evidence-based synthesis. Every module follows a four-stage shape - psychoeducation, exercise, practice, reflection - so your client always knows where they are in the arc, and you can predict the homework rhythm without inspecting every lesson.

The library is curated against current evidence, not just popular CBT one-liners. You'll find specialised tracks for OCD (ERP-structured), trauma (PE / CPT-aligned), eating disorders (CBT-E), perinatal, grief, identity formation, attachment-informed work, and DBT skills groups.

Yes. Library → Custom → New Module. You write your own in the same four-stage shape, save it to your private library (or shared with your practice's team), and assign it like any preset. Anchr Assistant can suggest key techniques and pull in evidence citations to enrich the module if you want.

Your custom modules are visible only to you (or to the team you nominate) - they're not added to the public library and they're never used to train external models.

Every lesson has its own private note space on your client's phone. They jot what landed, what bounced, what a concept reminded them of, what they couldn't shake - captured the moment they had the reaction, not three weeks later when they're trying to remember.

Those notes flow back to your portal in real time, beside the lesson they belong to. So Tuesday morning when you sit down to prep, you don't ask "how did the catastrophising worksheet go?" - you open their lesson note. "This one hit hard - the part about Sunday nights before work sounded exactly like me." You walk in already two layers deeper than the last session ended.

You curate - pick 2-6 modules from the library or your custom set that fit the formulation. Anchr ladders the assignments across the gap between sessions, paced so they're not overwhelmed and not idle. Your client opens the app and finds exactly what you wanted them doing, already sequenced and scaffolded.

If they get stuck on a tricky concept at 11pm, Talk it Through is there - module-aware, walking them through it in their own data. Every completion, skip, and lesson note flows back to your portal so you can see, before next session, what landed and what didn't.

Yes - one Anchr, your client uses it on their own or alongside you. When you assign a module, that becomes their guided track. When they're not in your module, they can still use Anchr Guide and Talk it Through in self-help mode and explore the library freely.

What changes for therapist-managed clients: certain consumer-side gating (the "refine pathway every 4 weeks", the explore-2-modules-first gate, the auto-reassessment cadence) is switched off - you're driving the pathway. The Living Compass and mood / defusion features remain fully theirs.

The full suite most AU psychologists use weekly:

• PHQ-9 (depression) and PHQ-A (adolescent variant, ages 11-17)
• GAD-7 (anxiety)
• K-10 (psychological distress)
• WSAS (work & social adjustment)
• DASS-21 (depression, anxiety, stress subscales)
• ORS / SRS (session-by-session outcome and alliance - Lambert tradition)
• PCL-5 (PTSD)
• AUDIT-C (alcohol use screen)
• C-SSRS (suicide severity - see Safety & Risk below)

Instruments can be sent to your client between sessions (push notification + magic link) or administered in the room on paper / on screen. Either way, the score, item-level responses, and trajectory land in the chart.

MBC is the practice of using validated instruments at every session (or every few sessions) to track change, not just at intake and termination. Anchr's MBC engine computes trajectory bands for each instrument using validated Reliable Change Index (RCI) and Clinically Significant Change (CSC) thresholds - so a 4-point drop in PHQ-9 isn't just "looks like an improvement", it's coded as reliable improvement, not reliable, or deterioration against published norms.

The engine surfaces divergence signals automatically - if mood, homework, and screening trajectories suggest something different from your last formulation, it flags before next session.

Yes. From the screening tab, pick the instrument and choose Administer in-session. You can hand your client a tablet, score on paper and key in the item responses, or capture verbally during the session - either way it lands in the same chart as their between-session screens and feeds the same MBC engine.

The system distinguishes paper-administered from self-administered for AHPRA documentation purposes.

For 11-17 year olds, Anchr branches automatically to age-appropriate instruments - PHQ-A for depression instead of PHQ-9, age-adjusted K-10 framing, and developmentally appropriate language across the safety screening flow. The MBC engine has separate RCI/CSC thresholds for the pediatric instruments.

Adult instrument variants (PHQ-9, GAD-7, etc.) remain available if you prefer them clinically - the platform asks once and remembers your preference per client.

The Living Compass is your client's values map - an ACT-derived bulls-eye where they rate their alignment with each life domain (relationships, work, body, identity, etc.) and the importance they place on each. Drift between alignment and importance is the workable material.

You see it in their chart - it informs the formulation, the brainstorm pre-session pull, and Anchr Guide / Talk it Through's framing. It's their instrument, but you read it like a values genogram.

Five consent kinds, each backed by a versioned template and a signed record:

• Informed consent for treatment
• Telehealth consent (jurisdictional, video / audio modality)
• Recording & AI scribe consent (separable from telehealth - your client can do video but not be recorded)
• Cancellation / fees policy (the same policy that fires the cancellation-fee ledger)
• Financial / Medicare claiming consent

Templates are versioned. If you update your cancellation policy on March 1, every fee raised before that date references the prior version (with the prior policy text snapshotted onto the fee row); fees raised after reference the new one. Signatures are captured by typed name, stylus, verbal + therapist countersign, or scanned PDF upload - whichever fits the moment.

ROI is the third-party disclosure layer - your client's authorisation for you to release information to a named external provider (GP, psychiatrist, school, employer, NDIS coordinator, insurer, family member). Under APP 5/6/12 of the Privacy Act 1988 this needs documented, scope-bounded, time-bounded consent that's distinct from consent for treatment.

Every ROI in Anchr is scope-tagged (which kinds of information - risk only? full notes? assessment results?), time-bounded (auto-expires unless renewed), revocable in one click, and writes a disclosure event audit log every time you actually exercise it - so if your client asks "who has seen my information?", you have a complete answer.

The chart export is an AHPRA-compliant PDF envelope that includes: demographics, all signed session notes and addendums, the case formulation, treatment plan, risk and safety plan, all screening assessment scores and trajectories, MHCP referral + session count, GP letters sent, ROI authorisations, consent records (with template version snapshots), and the full appointment history. Multi-attendee appointments are exported with all attendees correctly attributed.

Use it for transfer-of-care, subpoena response, AHPRA notification, audit, or your client's own access request. It's deterministic - same input, same output - so two requests at different times produce comparable files.

Yes. All clinical data is stored in the AU-Sydney region - the database (Supabase), the recording / transcription pipeline (Daily.co geo=au, Deepgram AU region), and all client / therapist file uploads. We don't replicate clinical data outside Australia.

The one current exception is AI inference - the LLM API call itself currently routes through Anthropic's US endpoint (no model training, zero data retention). We're actively migrating this to Amazon Bedrock in Sydney so the inference call never leaves AU; see the Roadmap section for state. Today, transcripts and prompts spend at most a few hundred milliseconds in the US inference path and are never persisted there.

Every PHI access event - every chart view, every note read, every appointment edit, every ROI exercise, every AI call - lands in a tamper-evident phi_audit_log table. Each entry captures who, what, when, the resource type, and a structured metadata payload. Cancellation-fee creations, multi-attendee appointment changes, and chart exports all fan out one audit row per attendee so co-attendees can each see the activity on their record.

Your clients see a redacted version of their audit log in their consumer-app Settings (the Therapist Access Log), so they always know who has touched their data. You see the full version in your own account-settings audit view. AHPRA / Privacy Commissioner audit requests are served from the same source.

Clinical records (notes, formulations, screening, safety plans) are retained for seven years after the last appointment - the standard AU clinical-record retention period - or longer if you specify (for pediatric clients, seven years after age 25, in line with State health-records legislation). Raw audio recordings are retained for 90 days by default then auto-purged; the structured transcript and extraction remain with the chart.

Account-deletion fires a comprehensive cascade: clinical PHI is removed, the audit log is anonymised (the event happened but the subject ID is stripped) per HIPAA-tradition retention discipline.

Yes - first-class. Every client chart has an MHCP referral tab where you capture the referring GP, referral date, Medicare item, eligibility category (standard Better Access 6+4, eating-disorder 40-session, chronic disease management, pregnancy support), session entitlement, and any associated documentation.

The platform tracks sessions used vs. remaining against the entitlement automatically as you create appointments, flags when you're approaching the review trigger (session 6→7), and tells you when an MHCP has expired or needs renewal. Multi-attendee sessions decrement correctly per attendee.

Standard Better Access is 6+4 (six initial, plus four after the review letter), reset annually on calendar year. Anchr tracks the count against the MHCP referral on each client's chart - cancelled appointments don't count, no-shows are configurable per your policy, multi-attendee appointments decrement per chart-owning attendee correctly.

The 40-session eating-disorder variant, 20-session perinatal, and CDMP plan variants all have their own counters and review triggers. The Assistant can also tell you across your whole caseload - "who's at session 6 and needs a review letter this fortnight?"

Five letter types, drafted from the chart already in front of you:

• Initial consultation letter (after sessions 1-2) - formulation, presenting concerns, planned approach.
• Review letter (sessions 6↔7) - the Medicare-critical one without which sessions 7-10 can't be claimed. Drafted from PHQ-9 / GAD-7 trajectory, MBC bands, formulation, and approaches actually used.
• Discharge summary - outcomes, final scores, recommendations for ongoing care.
• Referral back / onward - when you're handing the client to a psychiatrist or another psychologist.
• Progress report - for NDIS, insurance, schools, employers, or third-party requests.

Each is provenance-tagged (every claim citing the field it came from), drafted as a 30-second async job, ready for you to edit and sign. The single most-deferred admin task in AU private practice, lifted off your plate.

Yes - it's gated and surfaced. When a client hits their sixth Better Access session, the dashboard flags it; the Assistant proactively suggests drafting the review letter; the review letter draft template is structured around the elements Medicare actually expects (response to treatment, ongoing need, anticipated number of further sessions).

If you forget, the platform reminds you again at session 7. The letter doesn't have to be sent through Anchr - print it and post / fax / hand to the client, or email it to the GP through your existing channels. We just take the drafting load off.

From the schedule view, click any empty hour and pick Personal Event from the two-button chooser. Add a label (lunch, admin block, supervision, school pickup), pick recurrence if needed, save. Or just ask the Assistant: "clear my Friday afternoon" - it queues the personal event as a pending action, you confirm, done.

Personal events block availability so the appointment creation flow won't double-book you. They're private - clients don't see them.

Yes - the multi-attendee appointment spine is live. When creating an appointment, you can add up to 8 attendees in four roles: primary_client, secondary_client, collateral (parent, partner attending in support of the client of record), and support_person. Primary and secondary clients each get their own chart-side audit row when the appointment changes.

Two things are deferred for multi-attendee at the moment: (a) AI-drafted notes - couples / family sessions still need to be written by hand because the AI scribe is one-speaker-per-chart in the current release. (b) Per-attendee consent to record is still under development; recording in multi-attendee sessions defaults off until each chart-owning attendee has signed. See the Roadmap section.

You set the billing primary when you create the appointment - this is the chart that gets debited for the session count (against their MHCP if applicable) and against whom any cancellation fee gets raised. The default is the first attendee listed; you can override.

If neither partner is Medicare-billing the session, you can leave billing primary unset - the appointment still creates, but the session count doesn't decrement against an MHCP.

When you create an appointment, your client gets an email with a join link and an iCal attachment automatically - for both telehealth and in-person sessions. The push notification on the day links straight into the session.

You can configure reminder timing per your practice's defaults (e.g., 48h, 24h, 1h) and toggle reminders on / off per client. If you reschedule by more than two hours, the reminder cycle resets so they don't receive a stale 24h reminder for an appointment that moved.

Telehealth has a configurable 5-minute join grace window; the waiting-room indicator tells you when they join. If they don't show, you mark the appointment as no-show from the appointment row - this triggers your cancellation fee policy if you've configured one (see Billing below).

If they cancel inside your late-cancel window (24h or 48h depending on your config), marking it cancelled also triggers the fee. Outside the window, no fee. Your client sees the policy text snapshot on their consent page so the rules are visible upfront.

C-SSRS (Columbia Suicide Severity Rating Scale) is the validated screen for suicide risk. Anchr runs it as part of the standard screening cadence and treats it as a safety gate - if it screens positive, the platform will not let you sign a session note without a current safety plan attached.

That's by design. AHPRA expects a documented safety plan when active suicide risk is identified; Anchr makes the documentation the path of least resistance rather than an extra checkbox you might forget.

Standard Stanley-Brown structure: warning signs, internal coping strategies, social contacts and settings for distraction, people to ask for help, professional supports (you, GP, on-call psychiatrist), means restriction. Plus a therapist-only notes section for clinical reasoning that doesn't go to the client.

The client can see their plan from their app at any time - one tap from the home screen when they need it. Updates from either side (you edit, they edit) are versioned and timestamped; you always see the current plan and the change history.

Three things happen, in order:

• Anchr Guide / Talk it Through never silently filters crisis content. The conversation continues with appropriate clinical scaffolding, even at low model confidence. Filtering crisis language to protect the AI from awkwardness is how people slip through cracks.
• The crisis-resources surface appears in-conversation - Lifeline, 13YARN, Suicide Call Back Service, Beyond Blue - with one-tap dial.
• You see a divergence signal on your dashboard before the next session. The chart tags the conversation, surfaces it on the between-session activity feed, and (if your supervisor is configured) optionally notifies them too.

The AI never replaces your clinical judgement - it surfaces faster than you could otherwise see.

Yes - supervisor recovery is the platform's escalation surface. When you're on leave / unwell / unreachable, a nominated supervisor inherits read-access to flagged clients and can act in your stead (note-write is locked to you; clinical-decision authority stays with you).

Outside of leave, supervisors see what you nominate - per-client opt-in, with audit logging. C-SSRS-positive clients and clients with high-severity divergence signals can be auto-shared with supervisors if you set that as your default policy.

Monthly via Stripe. Card on file, charged on your billing anniversary. You manage payment methods, view invoices, and update billing details through the Stripe Customer Portal accessible from your account settings - everything PCI-compliant, nothing card-related touches Anchr's database.

Cancel any time. No annual lock-in, no setup fees, no per-session charges, no surprise add-ons. The price you sign up at is the price you pay.

Anchr - $149 / month per seat. The full platform: EHR, scheduling, telehealth, AI scribe, Anchr Assistant, the full module library, MHCP referrals, GP letter drafting, consent ledger, ROI, chart export, CPD log, supervision integration, direct client charging via Stripe Connect (0% Anchr take-rate), the works.

Founding 200 - $99 / month per seat, locked in for life. Same full platform; first 200 sign-ups only. We're using founding members to harden the product through real practice - you get the price, we get the feedback loop.

Group practice add-ons: additional seats (volume-discounted), supervisor seats (read-only, no charge for first one), annual prepay (15% off), group-of-three founding-tier bundle.

Yes. Anchr now bills clients through the platform end-to-end - card on file at intake, one-tap session charge from the appointment row, automatic cancellation-fee charging, refunds, receipts. Stripe Connect runs under the hood; Anchr takes 0% - the full session fee lands in your account on Stripe's standard payout schedule. The platform never touches the money.

Setting up. Connect your Stripe account from Settings → Billing → Setup - standard Stripe Connect OAuth, two clicks, takes about a minute. New Stripe account fine; we use Connect Standard so you retain full ownership of the connected account (chargebacks, disputes, payout schedule all yours). Once connected you set your session rates (per therapist, per modality, per duration - e.g. initial consultation, 60-min, 90-min couples) and your cancellation-fee policy.

Card on file. Send your client a magic-link card-capture page from their billing tab (or trigger it from the appointment row). They land on an unauthenticated Stripe Elements form, see your charge consent policy text, sign with typed name or stylus, and the card saves to a Stripe Customer scoped to (this client, this therapist). If a client sees two therapists in the same practice they have two separately-stored cards - clean separation, no cross-therapist leakage.

Charging a session. From the appointment row in the schedule, hit Charge session, pick the rate, confirm. The PaymentIntent fires off-session (no client interaction needed) against the saved card; the row shows live Stripe lifecycle status (processing → succeeded / failed / requires action). The rate at the time of charge is snapshotted onto the ledger row so historical receipts can't be poisoned by later rate edits. Refunds are one tap.

Audit-trail. Every charge, refund, consent capture, card add and card removal lands in the same phi_audit_log as the rest of the chart. Stripe webhooks reconcile state in both directions so the ledger and Stripe never drift.

Configure your policy in Settings → Cancellation Fees: late-cancel window (6-168 hours), no-show window, fee amount (flat cents or % of session fee), and the policy text your client sees on their consent page. When a session is cancelled inside the window or marked no-show, the trigger automatically raises an auditable fee row against the billing primary, with the policy text snapshotted onto that row for defensibility if your client later disputes.

The fee row goes into your Cancellation Fee ledger (separate inbox in Billing). Status lifecycle: pending → invoiced / charged → paid | waived. You can waive at one tap with a reason; you can mark paid manually with the invoice number captured for reconciliation if you invoiced outside Anchr.

Pick your charge method per rule. Each cancellation-fee rule now has a default_charge_method: manual_invoice (the original behaviour - raise a ledger row, you invoice externally) or auto_stripe (charge the saved card automatically when the fee fires). Auto-Stripe needs the client to have a card on file at the time of cancellation; if there's no card, the row falls back to pending-invoice and prompts you to send a payment-setup link.

Yes - this is now live. Set the default charge method on each cancellation-fee rule to auto_stripe, and when a session is cancelled inside your late-cancel window or marked no-show, the linked Stripe PaymentIntent fires automatically against the client's saved card. The fee row tracks Stripe lifecycle state in both directions: when the charge succeeds, the row flips to paid; if it fails (insufficient funds, card declined, expired), the row stays open and surfaces a retry-charge action plus a last_charge_failure_at timestamp so dunning is visible at a glance.

If the client has no card on file at the time the fee fires, the auto_stripe rule falls back to pending invoice and you can send a payment-setup link from the fee row to capture a card before retrying. You can also override per-rule (e.g. keep low-friction Medicare-eligible clients on manual_invoice, switch full-fee private clients to auto_stripe).

Stripe Connect Standard handles the dispute and chargeback path through your own Stripe dashboard - you retain full ownership of the connected account. Anchr's role is to fire the charge against your policy and reconcile the ledger; the financial relationship is between you and Stripe.

Not in v1. Anchr captures the appointment billing metadata (MBS item, bulk-bill flag, funding source, gap fee) and tracks the MHCP session count against the entitlement - but direct claim transmission to Medicare Online is still on the roadmap. For now, you export the appointment data and submit through your existing pipeline (Tyro Health Online, Halaxy Claims, Medipass, etc.) - or for solo practitioners, the manual Medicare PRODA flow.

Direct client charging via Stripe Connect (session fees + cancellation-fee auto-charge) is already live - Medicare claims submission is the next major billing release.

Yes - $99/month, locked in for life, for the first 200 sign-ups. Same full platform, same future features (including all paid roadmap items when they ship). No price increases for founding members ever. We're using the founding cohort to harden the product against real practice variety; in exchange you lock in the price.

Yes. The CPD log knows the AHPRA Psychology Board cycle (1 December → 30 November), the 30-hour annual requirement, and the ≥10 hours of peer / individual supervision minimum. Add an entry from the CPD page or via Anchr Assistant ("log 2 hours supervision with Dr X on Wednesday"), and it's allocated to the current registration year, with a running balance against the target.

At registration time you export the CPD ledger as a single PDF (or CSV for spreadsheet folks) - that's the audit document AHPRA wants to see if you're randomly audited. The framework supports PACFA, AASW, and ACA cycles too - pick your governing body at setup.

Six categories matching the AHPRA framework: supervision (peer or individual, with hours-toward-minimum tracking), training (workshops, courses, certificates), conferences, reading (journal articles, books, papers - your choice on what counts), teaching (presenting at a workshop, lecturing, tutoring), and other.

Attach evidence (PDF certificate, receipt, abstract) to each entry so the audit trail is complete. The supervision sub-category surfaces toward the 10-hour minimum; everything else counts toward the 30-hour total.

Highest priority slices already in development:

• Multi-attendee AI scribe - couples / family sessions get the same AI note-drafting and extraction pipeline as 1:1, with per-attendee speaker attribution and per-attendee consent surfaces.
• AU-resident AI inference via Amazon Bedrock (Sydney) - so the LLM call never leaves AU.
• Medicare claims submission - direct integration with Medicare Online so claims fire from the appointment row.
• Multi-clinic / org admin - clinic-level reporting, shared cancellation-fee ledger, group billing.
• Custom video SDK - in-portal video (instead of the current Daily prebuilt iframe) so the call doesn't leave the chart context.

Recently shipped: Direct client charging via Stripe Connect - card on file at intake, one-tap session charge from the appointment row, auto-charge for cancellation fees, refunds, receipts. 0% Anchr take-rate.

Founding members get input on prioritisation; we share the live roadmap with subscribers.

Book a guided onboarding call from the landing page. We bring your caseload across via CSV import (name, email, DOB, basic contact, MHCP if known), get your cancellation policy and consent templates configured, walk through your first telehealth and first AI-drafted note in real client work, and stay on for the trial fortnight to answer questions.

For appointment history we don't try to back-migrate - you keep your old EHR as read-only reference for archival, and start new appointments in Anchr. Most therapists find the cleanest cutover is end-of-month: archive the old system, run Anchr from the 1st.

In-app help inside Anchr Assistant - ask "how do I do X?" and you'll get the answer plus a deep-link to the right page. Email support at therapists@anchr.health - you'll hear back within 1 business day. For founding members and Group plans, there's a private Slack with direct access to the team.

Book a demo for live walkthrough from the for-therapists landing page. We'll show you the full workflow against a realistic case - pre-session brainstorm → telehealth → AI note → MHCP review letter draft - in about 25 minutes.

From your account settings: Delete Account. This fires a comprehensive cascade across every table that holds clinical PHI - session notes, appointments, screening responses, safety plans, consent records, ROIs, MHCP referrals, CPD entries, custom modules, the lot. The phi_audit_log is anonymised (the event happened but the subject ID is stripped) per long-tradition HIPAA-style record-keeping discipline.

For a single client's deletion request (Privacy Act APP 13 access / correction): from their chart, Delete Client Data. The clinical-record retention obligation (seven years post-last-appointment for AU psychology) may mean some data is retained in encrypted-archived form - the export-then-delete flow gives them their full record before the cascade fires.

Subscribe to the founding-member newsletter from the landing page - monthly product updates, the deferred-features ship dates as we lock them in, and early-access slots when major new surfaces (multi-attendee AI, auto-charge, Bedrock AU inference) ship. No spam, no marketing fluff, just the actual product trajectory.