Practice Brain Research | Medical Advertising Agency

outer ring

Applications — the doors to the world

Door	Direction	Access
Google Business Profile	reviews in · posts + replies out	draft → approve
Instagram / Facebook	posts out · DMs in	scheduled, approved
Email (practice address)	inquiries in · sends out	draft → approve
Website / CMS	pages + gallery out	draft → edit
Phones	call recordings in — transcribed on this computer	read-only
RealSelf	answers + reviews	draft → approve
EMR / PMS	not connected, ever — other side of the building

Every door starts read-only. Write access is earned, one job at a time. Scoped keys, not promises.

the full connector catalog →next ring in: JOBS →

ring 3

Jobs — what runs on a rhythm, one tap

Every day	Produces
inbox · 9a + 2p	every new inquiry answered, drafted
reviews · morning	replies drafted + today's ask list
follow-ups · morning	every touch due today, drafted
Every week
social · Monday	6 posts drafted + scheduled
case study · new pair	carousel + gallery + caption
report · Friday	the one page
Every month
web + SEO pass	fixes listed and drafted
referrers + past patients	touches + offer batch

Background tools — no AI, never touched

call transcription (on this computer)	call themes, counted
asset intake	photos renamed + coded + indexed
review pull	new reviews filed daily
form-inquiry pull	every lead becomes a row with a next date

A call ends → the transcript lands → next week's FAQ writes itself. The folder fills itself; the jobs read what lands; every result is one dated ledger line.

$8–15K

per lost case — the follow-up sequence (day 2 / 7 / 21 / 45) runs on every unbooked consult

2 min

inquiry → drafted reply, including the third that arrive after hours

5 min

the doctor's entire marketing job: one page, Monday

Every routine a practice can run

Grouped by what starts them. The ones marked sold as a product are things vendors already package and charge for separately — proof they work, and a rough map of what you would otherwise be paying for piece by piece.

Every routine above is a job you already knew you wanted. The half almost nobody builds is a job you did not — a standing question the brain researches on a schedule →

← outer ring: APPLICATIONS next ring in: MEMORY →

ring 2

Memory — everything the practice knows

Eleven numbered folders. Plain text. A human can read every file; so can the machine. It opens in Notepad in ten years.

Everything a practice owns, and where it goes

Most practices have far more than they think — scattered across three laptops, a photographer's Dropbox, an old agency's account, and someone's phone. This is the sweep. Tick what you have; the gaps are your first month of work.

Click any card to mark it — the count updates as you go

1 yr

of content in the camera roll — pairs become case studies within 7 days of landing in 02-ASSETS

20 Qs

every caller asks — counted in 07-AUDIENCE, published as FAQ, pages, scripts

0 names

topics, not people — no patient identity anywhere in the tree

Every file on the clickable map — open any of them: the map ↗

← ring 3: JOBS next ring in: SKILLS →

ring 1

Skills — one card, one job, done the same way every time

Each is a one-page playbook in SKILLS/. Written once, followed every time. A new coordinator reads them — that is the training.

reads what it opens first makes what lands on your screen A0 finishes alone A1 one-word approve A2 you edit

The full sweep — every skill worth having

You will not build all of these, and you should not try. Build the four above first. This is the map of what else is possible once those are boring and automatic.

← ring 2: MEMORY the center: CLAUDE.MD →

outer ring, in detail

Connectors — what plugs in, and how far in

What a connector actually is

A door between the AI and one of your accounts. That is the whole idea.

Without one, you are the connector. A review comes in, you open the tab, copy the text, paste it into the AI, read the reply, copy it back, paste it into Google. Six steps, one review, and it only happens when someone remembers.

With one, the AI opens the reviews itself, drafts every reply against your rules file, and puts the approved ones back. Forty reviews takes the same effort as one.

Every connection starts read-only and earns write access one job at a time. Scoped keys, not promises. Four tiers below — and one that never connects at all.

Every tool a practice asks about

Checked one by one in August 2026, against each vendor's own documentation. Four tiers: connect it today, build against it today, apply and wait, or it does not exist. The last column is the part nobody tells you.

Tool	Tier	What you get	The catch

The rule that keeps it safe: the brain drafts; your systems send. It writes the words — your email, your scheduler, your list tool press the button.

← the ring view what runs on it: JOBS →

solutions by practice

Same skeleton. Different practice, different first move.

The rings don't change. What changes is where your money is leaking — so the build order changes. Pick the one you are.

Where these numbers come from
Every figure on these five cards is the sourced arithmetic on what it's worth — lost-case value, reply-time, review counts, staffing cost — each traced to a primary source and dated there. Nothing here is a projection.

the arithmetic behind each card → the 90-day build order → the files for your type → every source →

← the center how it's actually run →

the honest answer

Obsidian? Notion? Which app is the brain?

None of them. The brain is the folder. Everything else is a window you look at it through — and windows are replaceable, which is the entire point. A practice brain written in plain text can be opened by any of these tomorrow, or by whatever replaces them in five years.

THE LOCK-IN QUESTION, ANSWERED

Your files outlive whichever tool you started with

This was the reasonable objection a year ago. It has since been settled by an open standard: AGENTS.md — plain markdown, no schema, placed in your folder. Over 60,000 projects use it, and it is now stewarded by the Agentic AI Foundation under the Linux Foundation rather than owned by any vendor.

Tools that read it today include Codex, Cursor, GitHub Copilot, VS Code, Zed, Warp, Devin, Windsurf, Jules, Cline, Roo Code, opencode, Amp, Factory, goose, Kilo Code, Junie and Augment. Claude Code uses its own CLAUDE.md and reads the other by import or symlink — a one-line file, not a migration.

Practical meaning for a practice: if you switch tools in two years, you copy a folder. There is no export, no vendor to ask, and nothing to lose.

the standard ↗how Claude reads it ↗

If you want to see the Obsidian version

Two of the most-watched builds in this space pair an Obsidian vault with Claude Code. Same markdown files underneath — Obsidian is just the window.

Brains actually running in production

Not concepts — public, dated, verifiable builds by people who publish their numbers.

Who teaches this, and where

The people publishing real builds right now. Subscriber and view counts checked August 2026 — this is a crowded, fast-moving field, which is itself the signal.

The open-source landscape

Star counts pulled from GitHub in August 2026. You do not need any of these to build a practice brain — but they are where the ideas come from, and several are worth an afternoon.

Everyone above teaches the same two things: store your notes, automate what you already do. Nobody is building the third →

← solutions by practice resources + videos →

learn it yourself

Resources, ring by ring

Everything here is free and real — official documentation, open repositories, and the videos that teach this well. Organized by which ring it builds.

References — every claim on this site

Nothing here is asserted from memory. Each number below traces to a primary source, fetched and dated. Where a figure comes from a vendor's own marketing, it says so.

← how it's run see it as slides →

what it is worth

The arithmetic

Run your own numbers into these. Every figure below is either a practice's own arithmetic or a sourced industry range — nothing is a projection dressed up as a promise.

The hours, per year

Not eliminated — moved. The work still happens; it stops happening in a person's evening.

The job	Done by hand	With the brain	Back per year

The moves nobody makes

This is the part that has nothing to do with AI. These are connections sitting in plain sight in every practice — the brain just makes them cheap enough to actually run.

The people question

This does not replace anyone. Say that plainly, because the room will be thinking it. What it does is let one coordinator produce what a marketing coordinator, a social media manager and a content writer would have produced between them — and stop the work from stopping when that person takes a holiday.

Role	Typical pay	Source

And the number nobody budgets for: US employers pay roughly 43% on top of salary in benefits and taxes — the March 2026 federal figure puts benefits at 30.1% of total compensation. A $60,000 coordinator costs about $86,000.

What the stack costs, and what this costs

Be fair about this comparison. A phone system is a phone system — the brain does not replace your lines, your scheduler or your practice management software, and it should not try. What it replaces is the layer above them: the writing, the timing, the judgment about what to say and when. That layer is what most of these platforms actually charge for.

Platform	Published price	Basis

Eight of the ten would not publish a price at all. Two have no pricing page on their own website. Where a figure exists it is either a "starting from" or comes from a third-party review site — and two of those third-party sources contradict each other on the same vendor's tier names. That opacity is itself worth noticing.

The graveyard

The argument for keeping your practice in plain files is not philosophical. It is this list — every one verified in August 2026. These are products people trusted with everything they knew.

Of roughly two dozen note apps checked, three keep your notes as plain files

Obsidian, Logseq and Basic Memory. Everything else stores your thinking in an encrypted vault, a SQLite database, a binary block store, or a cloud account with nothing on your disk at all. That is fine right up until the day it is not.

A practice brain is markdown in a folder specifically so that no announcement, acquisition or pricing change can take it from you.

← the 90-day guide get the files →

what you already own

Assets, and what each one turns into

An asset sitting in a folder is worth nothing. The value is in what consumes it. Every card below shows the same four things: what it is, where it lives, which jobs read it, and what comes out the other side.

Read one card and the whole system makes sense — because this is the actual wiring. Assets feed skills, skills feed routines, routines produce the work.

lives in the folder it belongs to read by the playbooks that use it becomes what it produces

One asset, six outputs

The clearest illustration of why this is worth doing. A single consented before-and-after pair — one photograph, already taken, already paid for — feeds six different pieces of work without anyone starting from scratch.

The sequences these assets run inside

Assets on their own are inventory. Assets inside a sequence are marketing. Here is what fires, in order, with what it needs.

the jobs that read these → the skills that turn them into work → the folders they live in → the naming rule, as a file → see them on the map ↗

← memory intake →

the library

Every source, in one place

Every source cited anywhere on this site, in one list — 91 of them. Official documentation, videos with their real view counts, open-source repositories with their star counts, vendor API docs and pricing pages, the salary and cost figures behind the arithmetic, and the dead-project evidence. Search it, filter it, or just scroll.

← curated shortlist the 90-day guide →

the working parts — what it does for you

42 tools, by what they're worth

Every row is a problem the owner feels on Monday morning, the plain-named routine that answers it, what lands on the screen, and the benefit. All 42 run on the same folder: the instruction page, the memory, a work folder. Every output is a draft a person approves. Marketing files only; nothing touches the chart.

Nobody builds 42. Build the review reply (#13), the consult follow-up (#1), the before/after slides (#12) and the Friday page (#24) first. The rest are a map for month four.

make money

save time

save money

better social

do what you love

— consults that book, patients who come back

#	The problem Monday morning	Tool	What lands on the screen	Benefit
1	Eleven people said "let me think about it" and nobody followed up	Unbooked-consult follow-up	day 2 / 7 / 21 / 45 drafts in the surgeon's voice, per consult, from a CSV	one recovered case a quarter pays for the year
2	Leads arrive at 9pm and get answered at 10am	Inquiry reply, twice a day	every form/email inquiry answered in draft, with the right page link and next step	the third of leads that arrive after hours stop going cold
3	The phone rang, nobody answered, nobody called back	Missed-call callback list	yesterday's unanswered + no-callback calls with a drafted text for each	the cheapest new patients you already paid for
4	Big treatment plan discussed, never heard from again	Quoted-and-quiet list	consults over $X with no booking after 30 days, one draft each	the money already sitting in your consult notes
5	Past patients who'd come back if asked	Reactivation batch (marketing list only)	quarterly note to patients on the marketing list, by procedure interest	repeat revenue without a campaign
6	Financing question kills the consult	Financing explainer	one page + one email in plain words, per procedure price band	fewer "I'll think about it"
7	Cancellations leave a hole in Thursday	Waitlist fill	drafted text to the three people who asked for sooner	the hole filled by lunch
8	Referring doctors forget you exist	Referrer letter, quarterly	a one-page update per referring office: new procedures, a case, a thank you	referrals that keep coming
9	Seasonal demand arrives and you're late	Season planner	the year's peaks by procedure (wedding, summer, holiday, tax refund) with a campaign draft 8 weeks ahead	ready before the rush, not during
10	Nobody knows which ad money worked	Lead-source report	every lead traced to its source, monthly; cost per booked consult, not per click	cut what doesn't book

2. Save time — the jobs nobody wants

#	The problem	Tool	What lands on the screen	Benefit
11	Six posts a week, every week, forever	Social calendar, monthly	30 days of posts drafted from the asset library and the season planner, scheduled, in your voice	the coordinator's Sunday night back
12	A great before/after pair sits in a camera roll	Before/after slide creator	pair → framed slide, carousel, gallery tile, caption, alt text; consent checked from the index	a year of content already shot, now out
13	Reviews pile up unanswered	Review reply	every new review answered in draft against the rules	40 reviews = the effort of 1
14	Same twenty questions on every call	FAQ from calls	the questions patients actually ask, counted from transcripts, answered once, published as FAQ + phone script	the front desk answers from a page, not memory
15	Tomorrow's consults, cold	Consult-day prep sheet	for each consult: what they asked on the phone, which page they read, the case to show, the objection to expect	the doctor walks in warm
16	New hire, no training	Training from the folder	a 5-day onboarding built from the instruction page, playbooks and last month's work	trained by Wednesday
17	Newsletter never goes out	Monthly newsletter	drafted from the month's case studies, reviews, and one thing worth knowing	a newsletter that ships on the 1st
18	Patient education handouts are 2019	Handout refresh	pre-op, post-op and procedure handouts rewritten in the voice, one per procedure	current, consistent, printable
19	Google Business Profile hasn't been touched since spring	GBP weekly	one post, Q&A answers, photo adds, hours check	the map listing looks alive
20	RealSelf and Google questions unanswered	Q&A answers	public questions answered in draft, topics not people	visible expertise, weekly
21	Video needs a script and nobody writes it	Shot list + script	for each procedure page: a 60-second script, shot list, and caption set	the phone comes out with a plan
22	Meet-the-team page is three people out of date	Staff bio writer	bios from a 6-question form, in the voice, with photo spec	done in a lunch break
23	Monday meeting has no agenda	Standup page	last week's ledger turned into: what ran, what's due, what's stuck	a 10-minute meeting
24	The doctor asks "how are we doing"	The Friday page	one page from the practice's own exports: reviews, calls, consults, follow-ups, ads	five minutes, Monday, done

3. Save money — replace what you rent, catch what you're overpaying

#	The problem	Tool	What lands on the screen	Benefit
25	Call scoring costs $150–500/mo and hides the report	Call review on your own machine	a month of calls transcribed locally, names stripped, read at once; report by what happened	$0, and the audio never leaves the building
26	The ad dashboard says $81 a lead; the phone says otherwise	Ad account audit	conversions re-counted against real calls and forms; true cost per booked consult	the $4,000/month that produced four calls, found
27	Six subscriptions, each knowing one thing	Subscription audit	every marketing tool, what it costs, what the folder now does instead	cancel what the folder replaced
28	Agency invoices nobody reads	Vendor report check	the agency's monthly report re-read against your own numbers; questions drafted	pay for results, not PDFs
29	Competitors change prices and you find out from a patient	Competitor watch, monthly	5–10 rival sites diffed: prices, new procedures, new providers, review counts	know first
30	Website has pages nobody visits and errors nobody sees	Site audit, monthly	broken links, slow pages, thin pages, missing schema, with the fix drafted	a site that keeps working

4. Better social — the practice looks like the practice

#	The problem	Tool	What lands on the screen	Benefit
31	Posts sound like a stock account	Voice-checked captions	every caption rewritten against the voice paragraph before it ships	it sounds like you
32	No idea what to post	Content from calls	this month's most-asked question becomes a post, a reel script and a page	content patients asked for
33	Reels need hooks	Reel script pack	ten 15-second scripts per procedure from the FAQ and case studies	the phone comes out with ten ideas
34	Testimonials never get filmed	Story interview kit	who to ask this month, the ask text, the 8 questions, the release, the shot plan	one real patient story a month
35	Case studies read like a chart	Case-study writer	pair + procedure + one line from the consult → a story in the voice, 150 words, three formats	pictures with a reason
36	Open house / event, no kit	Event kit	invite, landing page, three posts, two emails, the follow-up	one afternoon to build
37	Press and awards go unused	Press page + one-pager	every mention and award on one page, plus a media kit for the doctor	credibility that's findable
38	The doctor gets asked to speak and has nothing	Doctor speaker kit	bio (3 lengths), headshot spec, 3 talk abstracts, a slide deck from the case library	yes to the invitation

5. Do what you love — the doctor out of the marketing chair

#	The problem	Tool	What lands on the screen	Benefit
39	The doctor approves everything	Approval tiers	A0 runs alone (scheduling, filing), A1 one-word approve (replies, posts), A2 the doctor edits (case studies, claims)	the doctor sees 5 things a week, not 50
40	Nobody knows if the system ran	The ledger	one dated line per job: what came in, what went out, who approved	proof it ran while you were operating
41	What's changing in the field and nobody's watching	The standing question	a question the brain researches monthly ("what are patients asking about X", "who's new in our market")	you find the next thing before it finds you
42	You know what you want to say and never have time	Voice memo → everything	a 2-minute voice note becomes a post, an email, a page draft, in the voice	the doctor talks; the folder writes

Same list, law firm

Unsigned-consult follow-up · intake-call review on your machine · quoted-and-quiet matters · referral-source letters · case-result library → case study · practice-area FAQ from intake calls · GBP weekly · review reply · CLE/speaking kit for the partner · the Friday page.

What this is not

Nothing here reads the chart, the EMR, or the case file. Marketing files, marketing lists, consented photos, call transcripts with names stripped.
Nothing here sends. The folder drafts; your email, scheduler and text tool press the button.
Nobody builds 42. Build #13, #1, #12, #24 first. The rest are the map.

Each tool is one playbook file in SKILLS/. The market these replace is on the market; how they stay running is on jobs that run.

← the jobs ring jobs that run →

the case — what practices buy today

The market: what's sold and what it costs

Researched 7 Sep 2026, every row with its source. A single-location practice that buys the usual stack pays $680–$1,400 a month before overages, most of it on annual contracts, and every tool knows exactly one thing about the practice.

$680–1,400

per month, the bought stack, single location

2 of 11

healthcare voice-AI vendors that publish a price

$20–75

per user per month for a "HIPAA-safe ChatGPT" — all hosted chat

Scope: tools a 1–5 provider elective practice (plastics, derm, med spa, dental, ophtho) or a 1–10 lawyer firm buys or builds. Prices = vendor page unless marked "3P" (third-party reported, unverified). "Quote-only" = vendor publishes no price.

1a. Call intelligence — recording + transcription + scoring + missed-opportunity reports

Product	What it does	Published price	Source
Patient Prism	Dental/medical call tracking; AI + human-assisted scoring, booking verification, "missed opportunity" callbacks, staff attribution; add-ons: Voice AI agent, SMS AI, chat, reviews; 90-day pilot	Quote-only (SMB / Enterprise tiers). 3P: from $99/mo	https://www.patientprism.com/pricing/ ; 3P https://softwarefinder.com/emr-software/patient-prism
CallRail Conversation Intelligence	Call tracking + AI transcripts, summaries, sentiment, keyword spotting, lead qualification	Lead Tracking $50; LT Complete $95; Lead Conversion $150; LC Complete $195 /mo. CI = the two $150/$195 plans. 5 numbers + 250 local min included; overage ~$0.05/min, analysis ~$0.04/min (3P)	https://www.callrail.com/pricing ; 3P https://calltracker.io/blog/2026-08-08-callrail-pricing/
Weave Call Intelligence	Transcription, categorization, sentiment, AI task generation, "Scheduling Opportunity Detection" (scheduled/unscheduled per call), multi-location dashboards. Launched Oct 2024; enhancements Apr 2026	Weave "from $199/mo"; Pro/Elite/Ultimate quote-only; Call Intelligence = Ultimate tier only. 3P: Pro $249/mo, small practices $300–500/mo	https://www.getweave.com/call-intelligence/ ; https://www.getweave.com/pricing/ ; press https://investors.getweave.com/news-events/press-releases/detail/192/weave-launches-enhanced-ai-powered-call-intelligence-for-healthcare-practices ; 3P https://www.g2.com/products/weave-weave/pricing
Liine	Healthcare-trained AI call tracking: classifies call type, new-patient lead y/n, booked y/n, why-not; web-form callback automation; Google Ads integration; BAA included	Quote-only ("based on usage")	https://www.liine.com/pricing/
Call Box (medical)	Call tracking + cloud phones (GoTo Connect); human+ML review of every call; AI Mystery Shop Scorecard; source attribution	Quote-only	https://www.callbox.com/medical-practices/ ; https://www.capterra.com/p/136167/Call-Box/
Podium (AI Employee)	Reviews + messaging + phones; "AI Employee" answers leads/books; add-on to any plan	Quote-only on site. 3P: Core $399, Pro $599, Enterprise $999+/mo; AI Employee $99–399/mo; +$30/user phone seats	https://www.podium.com/pricing ; 3P https://checkthat.ai/brands/podium/pricing

Convoso: outbound call-center dialer, not a practice tool — excluded.

1b. AI receptionists / phone agents

Product	Vertical	Published price	Source
Arini (YC W24)	Dental; answers, books, reactivates; Dentrix/Eaglesoft/Open Dental write-back; "Watchtower" analytics; claims 1,000,000 booked appointments	Quote-only, per-location, segmented 1–20 / 20–100 / 100+ locations. 3P est. ~$249/mo	https://www.arini.ai/ ; 3P https://ainora.lt/blog/arini-ai-dental-receptionist-review-alternatives-2026
Annie (Dental Intelligence)	Dental voice AI; direct PMS scheduling; claims up to 70% of calls handled w/o staff, 35 hrs/mo saved, "3X ROI guarantee"	Quote-only	https://www.dentaleconomics.com/practice/article/55233048/meet-annie-ai-your-front-office-staffs-new-best-friend ; https://orthia.io/blog/dental-ai-receptionist-competition
Assort Health	Specialty-specific voice agents; $120M Series C (2026)	Quote-only. 3P: from $1,500/mo; Jul 2026 sales inquiry quoted $75k/yr minimum, 5-provider floor	https://www.fiercehealthcare.com/ai-and-machine-learning/assort-health-scores-120m-series-c-scale-voice-ai-agent-platform-healthcare ; 3P https://agentman.ai/blog/healthcare-voice-ai-pricing-comparison
Hello Patient	Voice + text + web chat, managed deployment, multi-location groups; $22.5M raise	Quote-only	https://www.fiercehealthcare.com/health-tech/hello-patient-secures-225m-investors-bet-ai-voice-agent-growth ; https://telnyx.com/resources/11-best-voice-ai-agents-healthcare
Simbo AI (SimboConnect)	Medical front-office AI phone agent	Official pricing page 404 (2026-09-07). 3P: Starter $95/mo 50 calls, Basic $270/150, Pro $800/500	https://simbo.ai/products/simboconnect/ ; 3P https://www.insighthealth.ai/blog/best-medical-ai-phone-reviews
Smith.ai	Law-firm-heavy; AI-first or human-first receptionist, per-call billing; Clio Manage/Grow, Lawmatics, Filevine sync	AI-first: Starter 30 calls $300/mo; Basic 90 calls $810; Pro 300 calls $2,100. Blog: AI ~$1–2/call, human ~$6–10/call	https://smith.ai/pricing ; https://smith.ai/blog/new-ai-receptionist-now-integrates-with-clio-manage-clio-grow
Ruby	Human receptionists (law firms core market); metered minutes	50 min $250; 100 min $395; 200 min $720; 350 min $1,210; 500 min $1,725 /mo	https://www.ruby.com/plans-and-pricing/
Zenoti AI Receptionist	Med spa/salon PMS; AI receptionist + text bot on "Hypergrowth" tier	Quote-only. 3P: $300–500+/mo/location, AI sold as add-ons	https://www.zenoti.com/pricing-zenoti ; 3P https://pabau.com/blog/zenoti-pricing/
Med-spa AI phone vendors (CallMyDoc, Brilo, Seqora, Ignitvio, Ringlyn, MedspAI)	Plastic-surgery/med-spa voice agents + "didn't book" nurture sequences; integrate Aesthetic Record, Boulevard, PatientNow, Symplast, Nextech	Quote-only (all)	https://www.callmydoc.com/plastic-surgery-phone-automation ; https://www.seqora.io/medspa ; https://www.brilo.ai/industry/plastic-surgery-clinics ; https://www.layer3labs.io/guides/ai-answering-service-for-medspas

Industry note (3P, 2026): of 11 healthcare voice-AI vendors surveyed, 2 publish pricing — https://telnyx.com/resources/11-best-voice-ai-agents-healthcare

2. Patient reactivation / recall / follow-up

Product	What it does	Published price	Source
NexHealth	Online booking, SMS/email recall + reactivation campaigns; Summer 2026 release adds lapsed/churned "reactivation opportunity" scoring	"from $299/mo" (3P: $350); Starter/Standard/Pro quote-only; 3P add-ons: recall/marketing campaigns $1,200–9,000/yr	https://www.nexhealth.com/resources/summer-release-2026 ; 3P https://www.themolarreport.com/learn/nexhealth-pricing
Weave	Reminders, 2-way text, recall, reviews, payments, phones	from $199/mo (see §1a)	https://www.getweave.com/pricing/
Solutionreach	Reminders, recall/recare campaigns, 2-way text	Quote-only. 3P: $400–800/mo (3–8 providers) + $500–2,500 implementation, annual contract	https://noshowcost.com/tools/solutionreach-pricing
Klara	Patient messaging, per-provider	Quote-only, annual contract	https://emitrr.com/blog/klara-pricing/
RevenueWell / Adit / Dental Intelligence	Dental PMS-mined recall + reactivation sequences	Quote-only	https://www.dentalbase.ai/blogs/practice-management/ai-dental-patient-reactivation-complete-guide
PatientNow "AI Growth Engine"	Aesthetics PMS; finds interested-not-booked patients, drafts campaign email copy	Quote-only	https://www.patientnow.com/ai-growth-engine/

3. Review automation

Product	Published price	Source
Birdeye	Site: quote-only, priced by location count. 3P: Starter $299, Growth $349, Dominate $449 /location/mo; annual auto-renew contracts	https://birdeye.com/pricing/ ; 3P https://costbench.com/software/review-management/birdeye/
Podium	Quote-only; 3P Core $399/mo (see §1a)	https://www.podium.com/pricing
NiceJob	Reviews $75/mo; Pro $125/mo (AI review replies, referral campaigns)	https://contractortoolstack.com/software/nicejob/ (nicejob.com/pricing failed TLS fetch 2026-09-07)
Swell	Healthcare-focused reviews + surveys + texting; AI drafts every review response	Quote-only
Weave / NexHealth	Review requests bundled in base plans (§1a, §2)	see above

4. "HIPAA-compliant AI" — who sells it and how they say it

Vendor	Positioning (quoted)	Price	BAA scope	Source
BastionGPT	"Want to use ChatGPT, but work in healthcare? Meet BastionGPT"; "Private, healthcare-grade alternative to ChatGPT"; "Built for clinicians, not consumers"; "handles sensitive topics other AIs refuse"	Professional $20; Professional Plus $45; Ultra $75 /user/mo	"HIPAA BAA included on every plan… Including the free trial"; hosted on Azure; claims 10,000+ orgs	https://bastiongpt.com/
CompliantChatGPT	"a medical copilot built for the way clinicians actually work"; "tokenizes all 18 HIPAA identifiers before your data reaches any AI model"	Starter $19.99; Pro $35.99; Full $63.99 /user/mo	BAA on paid plans; GPT + Gemini models	https://compliantchatgpt.com/pricing
Hathr.AI	"HIPAA Compliant Claude AI"; AWS GovCloud, NIST 800-171	Webapp $47/user/mo; API from $45/mo	BAA on all plans; Claude models	https://www.hathr.ai/pricing
Doximity Ask (was Doximity GPT)	"Your Free, HIPAA-Compliant Workflow Assistant"; drafts prior auths, appeals, patient instructions; PeerCheck physician review	Free to verified US physicians/NPs/PAs/pharmacists/students	PHI allowed in prompts per FAQ	https://support.doximity.com/hc/en-us/articles/41850759500051-Doximity-GPT-FAQs ; https://www.doximity.com/
Anthropic (Claude)	BAA on HIPAA-ready services	Enterprise (quote) + 1P API usage	Covered: Claude Enterprise w/ HIPAA toggle (Chat, Projects, Artifacts, code exec, Voice, Web Search, Research, Skills); 1P Messages API; Claude Code CLI/desktop only with Zero Data Retention. NOT covered: Free/Pro/Max/Team, Batch/Files/Skills API, Computer Use, Web Fetch, Cowork, Design, Claude in Office, any MCP/connector/Chrome data sent to 3rd parties	https://support.claude.com/en/articles/8114513-business-associate-agreements-baa-for-commercial-customers ; https://support.claude.com/en/articles/15455031-covered-models-under-a-business-associate-agreement-baa
Microsoft Azure OpenAI	HIPAA-eligible under Online Services DPA; BAA default in EA/CSP	Usage	Text in/out covered; image/vision and realtime audio not covered (mid-2026, 3P)	https://www.aptible.com/hipaa-compliant-ai-tools/azure-openai-baa ; https://learn.microsoft.com/en-us/answers/questions/2265200/azure-ai-baa-hipaa-compliance
Google Vertex AI	Vertex covered under Google Cloud BAA; Gemini via Vertex inherits; consumer Gemini/AI Studio API not covered	Usage	3P summaries	https://aiprovidertrust.com/questions/hipaa-baa/ ; https://www.tactionsoft.com/ai-healthcare/baa-with-ai-providers/

Pattern: every "HIPAA AI" wrapper = (a) BAA on every tier incl. trial, (b) "your data never trains a model", (c) "alternative to ChatGPT", (d) $20–$75/user/mo. None sells a local-files/folder model; all are hosted chat.

5. "Second brain" / "AI operating system" / Claude-for-business creators

Creator / product	Model taught	Price / size	Source
Robyn "Jam" Anderson — Club Jam (Skool "jamoutai")	Daily micro-lessons, 30-day sprints, live guest sessions ("Zero to Claude"); "Official Claude Partner", Anthropic Claude SMB Trainer; SCORE contributor, Udemy	$47/mo; 447 members (Skool, 2026-09-07); site claims "400+ members", "7,500+ business owners trained", "93% stay"; 7-day trial. No folder/Projects curriculum named on site	https://www.skool.com/jamoutai/about ; https://www.clubjam.ai/ ; https://www.score.org/contributor/robyn-jam-anderson/
Early AI-dopters (Mark Kashef, Skool)	Claude Code course ("Claude Code Magic"), workflows	$77/mo; 1,300+ members (3P)	https://aifunnelinsider.com/early-ai-dopters-review-2026/
Claude Club (Skool)	Weekly prompt + workflow drops for content creators	$87/mo, 7-day trial (3P)	https://aifunnelinsider.com/claude-club-skool-review-2026/
Tiago Forte — "The AI Second Brain"	3-week live cohort, "personalized AI system that works like a new team member"; Cohort 2 Sep 17–Oct 8, 2026	Premium upgrade $3,500; base tier price not captured	https://www.buildingasecondbrain.com/ai-second-brain
MindStudio — "AI Operating System with Claude Code" (2026-05-26)	9-component architecture (identity, memory, tools, orchestration, RAG…) + "Four C's" (context/connections/capabilities/cadence)	Content marketing for MindStudio plugin; price not in article	https://www.mindstudio.ai/blog/what-is-ai-operating-system-build-with-claude-code
"Marketer's Guide to Claude Projects" (Substack, Aug 2026)	Projects as a "department or business folder": brand voice, ICP, positioning, templates in one place; Cowork reads a local folder	Free newsletter	https://getmarketingwithai.substack.com/p/the-marketers-guide-to-setting-up
Bear My Brand — "How Doctors Can Build a Personal AI Assistant Using Claude" (2026)	Claude Pro $20/mo + Projects for a medical assistant; Team $25/user for multi-doctor	Blog (agency lead-gen); page fetch 403	https://bearmybrand.com/blog/how-doctors-can-build-ai-assistant-claude-2026
Anthropic — Claude for Small Business (2026-05-13)	15 prebuilt workflows in Cowork (payroll planner, month-end close, invoice chaser, lead triager, content strategist, campaign runner…) inside QuickBooks/PayPal/HubSpot/Canva/DocuSign/Workspace/M365; "you approve before anything sends, posts, or pays"; free 10-city SMB Tour workshops May–Fall 2026	Included for paying subscribers (3P)	https://www.anthropic.com/news/claude-for-small-business ; https://beginnersinai.org/claude-smb-tour/

6. Law-firm equivalents (small firm)

Product	What it does	Published price	Source
Clio Duo / Manage AI	AI inside Clio Manage: drafts, summaries, billing entries, task extraction	3P: $39/user/mo add-on (Lawyerist); other 3P: $49–59; Duo no longer on pricing page, successor "Manage AI" quote-based (clio.com/pricing fetch 403)	https://lawyerist.com/reviews/artificial-intelligence-in-law-firms/clio-duo-review-artificial-intelligence-for-lawyers/ ; https://aitoolsbakery.com/blog/clio-pricing-2026/
Smith.ai	AI/human intake receptionist; Clio Grow/Manage, Lawmatics, Filevine sync	AI-first $300/30 calls, $810/90, $2,100/300 per mo	https://smith.ai/pricing
Ruby	Human receptionist + 24/7 chat for law firms	$250–$1,725/mo by minutes	https://www.ruby.com/plans-and-pricing/
Lawmatics	Legal CRM + intake + marketing automation + AI	3P: Lite $149/user/mo, 3-user min ($447 floor); Pro ~$199; Enterprise ~$299	https://softwarefinder.com/legal/lawmatics/pricing ; https://directory.lawnext.com/products/lawmatics/
CallRail (legal)	Same product as §1a; law is a named vertical	$50–$195/mo	https://www.callrail.com/pricing
Gavel (ex-Documate)	No-code document automation + client intake forms; Clio, DocuSign, Zapier	Lite $83; Standard $165; Pro $290; Scale $417+ /mo; annual = 2 months free	https://www.gavel.io/pricing
Spellbook	Contract drafting/review in Word	No public rates; 3P $99–199/user/mo, enterprise ~$350 w/ 6-mo min (late-2025 increase)	https://bindlegal.com/resources/comparisons/spellbook-pricing-2026/ ; https://www.aivortex.io/legal/compare/spellbook-pricing-2026/
Harvey	Enterprise legal AI	No public pricing; 3P $1,000–2,000/user/mo, 20–50 seat minimums, annual — not a small-firm product	https://thelegalprompts.com/blog/harvey-ai-pricing ; https://www.aivortex.io/legal/ai-tools/harvey-ai-pricing-2026/

7. Ambient clinical scribes (adjacent)

Product	Price	Source
Freed	Starter $39/mo (40 notes); Core $79; Premier $104 annual / $119 monthly; Groups custom	https://www.getfreed.ai/pricing
Heidi Health	Free tier; Clinician plan 3P USD 150/user/mo annual (Feb 2026 rebrand, up from ~$90); site shows tiers, no numbers	https://www.heidihealth.com/pricing ; 3P https://www.commure.com/blog-scribe/heidi-health-review
Abridge	Quote-only, enterprise; 3P $208–$800+/provider/mo; Best in KLAS ambient 2025 + 2026	https://www.veroscribe.com/blog/abridge-review-2026
Nuance DAX Copilot	3P $444–$600/provider/mo; Epic health systems	https://patientnotes.ai/compare/nuance-dax
BastionGPT AI Scribe	Bundled in $20/user/mo plan (§4)	https://bastiongpt.com/ai-medical-scribe

Price floor summary (single-location practice, 2026)

Stack a practice buys today	Monthly
Call tracking + AI (CallRail Lead Conversion)	$150
AI receptionist (Smith.ai Starter / Simbo 3P)	$95–$300
Reactivation + reminders (NexHealth)	$299+
Reviews (NiceJob → Birdeye)	$75–$449
HIPAA chat (BastionGPT/Hathr)	$20–$47/user
Scribe (Freed)	$39–$119
Total	$680–$1,400+/mo before overages, annual contracts on most

Sources for every figure are in the row that carries it.

← what it's worth the 42 tools →