The 9-Step Framework for Deploying Agentic AI in a Physiotherapy Clinic

Here is the part nobody says out loud to the physiotherapist who owns the clinic. Somewhere between your first few regulars and a full diary, you stopped being only the clinician and became the operating system.

Every new referral, every patient who needs to rebook, every "did we actually invoice that session", every patient quietly deciding whether to come back waits on your attention, which means the whole clinic runs exactly as fast as one owner-clinician can keep track of it between hands-on appointments.

This is the framework we use to take that operating-system job off the owner and give it to a system, without the risk that comes with letting software anywhere near patient records, clinical plans and money. It is nine steps in three phases.

Four before you build anything, three to build it, and two to do the thing the whole exercise is for, which is to remove yourself from the middle of every patient journey.

It is worth saying clearly what this is not. It is not a new clinic-management package, and it is emphatically not "let AI treat the patients".

The clinical judgement stays entirely with your physiotherapists. What the AI removes is the admin around the treatment, the referral filing, the record chasing, the rebooking, the follow-up nobody had time to make, the invoice that never got raised.

The assessment, the diagnosis, the hands-on plan and the progression stay exactly where they belong, with your qualified people. Nothing in these nine steps touches a clinical decision.

You quietly became the operating system

It happens slowly, and it happens because you were good at the job. In the early years you did the assessments, the treatment, the phone, the rebooking and the invoicing yourself, because there was nobody else and you could hold it all in your head between patients.

So the clinic grew. Then it kept growing, and the routing never left your head, because on a busy day it was always faster to just do the thing than to stop and write down how it should be done.

Now you are the one who notices a post-op knee patient has not been back in nine days, who remembers the shoulder referral that came in on Friday and still has not been triaged, who spots that a whole block of advisory calls and extra sessions never made it onto an invoice.

None of that is written down anywhere. It lives in your attention, in your memory, and in a WhatsApp thread with the front desk.

In a clinic, attention is the scarcest resource there is, because most of your day is spent with your hands on a patient and your phone in a drawer. Three things follow from being the operating system, and every one of them costs the clinic real money and real patients.

• The clinic runs at attention speed. An owner-clinician spending the gaps between patients chasing referrals, checking who has rebooked and working out who was billed is normal, and it is the most expensive triage in the building, because it is done by the person whose hands should be on a patient.
• Follow-up happens because somebody remembered. Which is fine until the week nobody did. A patient who finished a course of treatment and never got a check-in call does not complain. They simply do not come back, and they do not refer anyone, and you never find out why.
• Nothing survives your day off. A clinic where the owner is the router does not pause when the owner is treating back to back, or away for a course, or sick. Referrals pile up untriaged, patients lapse unrebooked, sessions go unbilled, and you discover it a fortnight later when the diary looks thin and you cannot explain it.

The instinct is to fix this by hiring another physio, or another receptionist. That works, and it also adds a salary, a training burden, and one more person who has to learn all the unwritten rules you were already the sole keeper of.

You have not removed the bottleneck. You have given it a second head to depend on, and the day either of you is off, the routing still stops.

The alternative is to write the routing down and let a system run the parts that never needed a clinical brain in the first place: the filing, the chasing, the rebooking prompts, the invoice drafts. That is exactly what the next nine steps do.

For a physiotherapy clinic, the patient experience is the marketing

One belief before the framework, because it decides how you read the rest of it. In a clinic, the patient experience is the marketing.

Not the website, not the sign outside, not the ad budget. The follow-up call that actually happened.

The rebook offered before the patient drifted. The check-in a week after discharge that made a person feel looked after rather than processed.

That is what turns one patient into three referrals, and it is produced almost entirely by the nine jobs we are about to hand to a system.

This is why the topic belongs on a growth marketing site and not in an IT catalogue. A finished patient who never got a follow-up is not an admin oversight, it is a referral you will never receive and a return course of treatment you will never bill.

A package that expired with three unused sessions inside it is not a scheduling slip, it is money the patient already paid you and goodwill you just spent. A session delivered and never invoiced is not a finance error, it is revenue you earned with your own hands and then gave away.

When the operating system is one tired owner-clinician, the patient experience frays in exactly the places a patient feels and a rival clinic asks about at the first appointment.

Which brings up the trap almost every growing clinic falls into. They try to grow by pouring more money into ads and more new patients through the door, into a clinic whose follow-up and rebooking still run through the owner between appointments.

That does not produce growth. It produces a fuller waiting room, a longer chasing list, more finished patients who quietly lapse, and an owner who is more exhausted, not more profitable.

This is the capacity problem, and it is one of exactly three things almost every stuck clinic is stuck on. The other two are getting the right patients in and converting the enquiry into a first appointment, and I have written the full diagnostic in the 3A Machine.

So the goal of deploying agentic AI in a clinic is not "use AI to do physio". It is to build the capacity that lets you grow without the care and the follow-up falling over.

Get the nine steps running and the same physiotherapists and the same front desk hold two or three times the caseload, with every finished patient followed up, every package watched, and not a single referral left sitting untriaged over a weekend. You fix capacity first, so that the patients you win actually stay and refer, and then you go and win more of them.

It is worth dwelling on one number here, because clinic owners underrate it every time. The cheapest patient you will ever treat is the one who already finished a course with you and had a good outcome.

Winning them back for a maintenance block or the next flare-up costs nothing but a well-timed message, and it is precisely that message, the follow-up almost no clinic reliably makes, that leaks first when the owner is the operating system. Fix that one leak and you have found growth you were already paying for and simply never collected.

The nine steps, in three phases

Here is the whole map on one page. Your clinic already runs all nine of these jobs today, whether or not anyone has ever named them.

The framework does not add work. It names the jobs, then decides one at a time whether a person does each one or a system does it.

The clinical jobs stay with people. The routing jobs move to the system.

Notice the shape. Four steps happen before anybody builds anything, and skipping them is the single most common reason a clinic's "let us try some AI" project quietly dies.

People buy the clever part first, point it at nothing in particular, and end up with a fast tool that does not know how the clinic actually runs, how a post-op patient differs from a walk-in, when a package is meant to be reviewed, which referrals need triage before booking. The order below is designed to stop exactly that.

Step 1. Diagnose: find what is actually bleeding

You cannot fix what you have not counted, and most clinics have never counted this. So the first step is a leak audit, and its only job is to put a number on the money and the patients already walking out of the clinic.

Not a survey of how everyone feels about being busy. A count of losses.

In a clinic the leaks are always in the same few places, and they are quiet, which is why they survive for years. Patients who finished a course of treatment, got a good outcome, and were never contacted again, so they neither returned for the next issue nor referred anyone, because nobody reminded them the clinic existed.

Packages bought in good faith with sessions still unused when they expired, which the patient half-remembers and quietly resents. Sessions genuinely delivered and never invoiced, an extra appointment squeezed in, a taping, a phone consultation, a bit of advice between blocks, because raising the invoice was a small annoying job nobody owned.

And the hours the front desk and the owner burn every single week chasing referrals, rebooking by phone tag, and rekeying a referral letter into the record by hand.

Add it up honestly and the total is almost always bigger than expected, and concentrated in two or three places rather than spread evenly. In most clinics the single biggest number is the finished patients who were never followed up, because each one is a lost return course of treatment plus the referrals they never made, and it compounds silently month after month.

That concentration is the gift. It tells you exactly where to point the build first, and it becomes the before number you use, at the very end, to prove the system paid for itself.

Step 2. Shadow: follow one patient from referral to discharge

Now you watch. Pick one live patient journey, ideally a post-operative case because it has the most moving parts, and follow it end to end, writing down every single thing that happens to it and every decision anybody makes.

From the referral letter arriving, through triage, the intake, the first assessment, the treatment block, the reviews, the rebooks, the package, the billing, right through to discharge and whatever follow-up does or does not happen after. Not the tidy version in a protocol document nobody opens.

The real one.

This is where you discover the clinic does not run on the written protocol. It runs on a hundred unwritten rules that live in your senior physiotherapists' heads and, more than anywhere else, in the messages between the front desk and the clinical team.

Which documents have to be on the record before a post-op patient can be booked in at all. How this GP always sends referrals and what is always missing from them.

The red-flag questions your lead physio always asks and never skips. The point in a knee rehab where a review is due whether or not the patient asks for one.

The rebook cadence that keeps a lower-back caseload progressing instead of drifting. The patients you never discharge without a maintenance conversation.

None of it is written down, all of it is load-bearing.

The excavation is the real work of this step. You read back through the threads and the notes and pull out the rules the clinic actually operates on, the ones that were never a formal decision, just a habit that turned out to be right and quietly became the standard of care.

That messy, lived-in reality is what you are about to encode. Skip it and you will automate the fantasy version of the clinic, ship it, and watch your physiotherapists quietly go back to doing it their own way because the system does not know what they know.

And be honest about the fragility this surfaces: in most clinics the real operating manual is one long-serving physio's memory and one long-serving receptionist's instinct, and it walks out of the door the day either of them leaves.

Step 3. SOP: write the intake and the follow-up down, and add a self-check

Everything you excavated now becomes written procedure. Plain language, step by step, in the order it really happens, with every rule and every exception stated.

This is the least glamorous step in the framework and it is the one that makes an agentic system possible, because a front-desk employee can only work a procedure that has actually been written. In a clinic there are three procedures worth writing first, and they map directly onto the three biggest leaks.

The first is the intake procedure. For a new post-operative patient it looks like a hard checklist: the GP referral, the operation note, a signed consent form, the red-flag screen, the baseline measures, and the agreed treatment plan, gathered in this order, with the job unable to advance to a booked first appointment until every item is present.

Written as a procedure, the messy reality becomes a gate a machine can enforce perfectly, every time, without a rushed front desk booking a complex patient in on a busy morning before the clinic actually knows what it is dealing with. Nobody with their hands on a patient has to remember it, because the gate remembers.

The second is the follow-up schedule, and this is the one clinics almost never write down, which is exactly why they almost never do it. When is the rebook prompt due after a first assessment.

When does a package get a review conversation rather than another silent session. When, after a patient finishes a course, does a check-in message go out, at one week, at six weeks, at six months, and what does it say.

Written down, the follow-up stops depending on somebody happening to remember a name, and becomes a schedule a system can run so that no finished patient is ever silently dropped again.

The third is a self-check, a simple triage the patient or the referrer can complete before they ever reach a clinician. A short set of questions that sorts a genuine musculoskeletal case that is ready to be booked from one that needs the red-flag screen or a GP conversation first, and that captures the basics so the assessment starts from something rather than a blank page.

It is not a diagnosis and it never pretends to be, the clinical judgement stays with your physiotherapists, it is a sorting step that stops your most expensive people spending their first ten minutes gathering information a form could have gathered.

Write all three for a smart new receptionist on their first day, not for a machine. If a sensible person could follow the procedure without asking a single question, a system can run it.

If it still needs somebody to "just know", the procedure is not finished, and you are not ready to build.

Step 4. Baseline: agree the numbers before you touch anything

The last step before the build, and it is thirty minutes that saves you a year of arguments. Write down the current numbers and get the owner and the lead physio to agree them.

How long it takes to file a referral to the right patient record today. What share of finished patients actually get a follow-up call.

How many packages expired last quarter with unused sessions still inside. How much delivered work is sitting unbilled right now.

How fast a new enquiry gets a first reply. Real numbers, honestly rounded, on the record.

You do this for one reason. In three months, when the system is running, memory rewrites history.

People forget how leaky it was, decide the follow-up was "always basically fine", and start to wonder what they are paying for. The baseline is the receipt.

It is what lets the weekly proof report at the end say "filing a referral went from about twenty minutes by hand to about two" and have that land as a fact both the owner and the lead physio agreed up front, rather than a vendor's claim.

It is also the moment to decide what "better" means for this clinic, so the build aims at a number and not a vibe. A clinic bleeding on finished patients who never come back is not chasing the same win as one drowning in untriaged referrals or one that keeps forgetting to invoice.

Name the number now. It is what everything you build next is pointed at, and it is the difference between a clinic that can say the system paid for itself in the first month and one that just feels a bit less frantic and cannot say why.

Step 5. The command center: the whole clinic on one screen

Now you build, and the first thing you build is the place the owner looks. One screen that shows the whole clinic at a glance, in the order that costs money and loses patients, so nobody reconstructs the state of the clinic from the diary, the notes system, the card machine and a group chat between appointments.

Build it on one spine, not twelve tools. This matters far more than it sounds.

Every job in the audit is tempting to solve with its own separate app, a rebooking reminder tool here, a review-request tool there, a document tool, a billing add-on, and a year later you have a dozen subscriptions that do not talk to each other and an owner who is now the integration layer between them, copying a name from one screen into another between patients. Build the whole thing on one foundation, one place the patient data and the clinic context live, and the next piece is nearly free, because everything it needs already exists on the same spine.

The centre of the screen is a patient radar: every active patient, ranked not by their next appointment date but by who is about to fall out of care. The post-op knee patient who has not rebooked in nine days sits at the top, in red, whether or not anybody thought to look for them, because a patient about to lapse is worth more attention than one who is already booked in for tomorrow.

Around it, the numbers the owner actually needs: how many patients need a call today, how many have not rebooked and are at risk of dropping off, how much delivered work is unbilled, how many packages expire this month.

Then the follow-up clocks. A clinic does not have one follow-up window, it has dozens running at once across the whole caseload, and a single shared diary flattens them into a comforting average that hides the one about to cost you a patient.

The command center pulls them apart and shows each on its own, so the rebook window on a lapsing patient, the expiry date on a half-used package, and the review due on an active plan are three separate clocks, and the dangerous one is the one you see first. Every clock is a patient you either keep or quietly lose depending on whether the window closes with a message inside it or empty.

Beside the clocks, the money on the floor: every session delivered, done, treatment given, and never invoiced. In most clinics this is a genuinely uncomfortable number the first time it appears on a screen, because the work was already paid for in the physio's time and effort and simply never got billed.

It was not a pricing decision. It was an extra session squeezed in, an assessment, a taping, a phone consult, and raising the invoice for it was a small annoying job nobody owned, which is a fair description of most of what leaks in a clinic.

And the document wall. An established clinic is sitting on an enormous pile of patient records, referral letters, operation notes, scan reports, consent forms, treatment plans, progress notes, tens or hundreds of thousands of files across folders nobody has fully mapped in years.

The command center reads that existing storage where it already lives, with no migration and no "please move every record into our new system", and makes it navigable and countable, so any document is found in seconds instead of a frightened hunt through folders while a patient waits on the treatment table and a physio's gloves come off to search.

One design choice worth naming, because clinic owners always ask, and rightly so in a healthcare setting. The command center is owner-locked, and the team gets read-only logins scoped to what they actually need.

The owner sees the whole board, the revenue, the caseload, the leaks. A physiotherapist sees the patients assigned to them.

The front desk sees the referrals and the rebooking list. Nobody can quietly change a number they should only be reading, patient records stay handled the way a healthcare setting demands, and the owner never loses the single honest view of the clinic that the whole build exists to give them.

Step 6. The procedure as data: the clinic's mind

The procedures you wrote in step 3 are documents, and a document just sits there. In step 6 you turn them into data the system can reason over: every protocol, every intake checklist, every discharge rule, every rebook cadence, every red-flag screen, and every good answer the clinic has ever given, connected by meaning rather than filed in folders.

This is the clinic's mind, and it is what makes the front-desk employee in the next step sound like your clinic instead of a generic model that has never met a post-op knee.

In practice it means anybody, the owner, a physio, a new receptionist, can ask a plain question and get the clinic's own answer, not the internet's. What is our intake checklist for a new post-operative patient.

What is the rebook cadence for a lower-back plan. What are the red flags we always screen before booking.

When do we send a discharge follow-up and what does it say. The answer comes back grounded in the clinic's own written procedure, with the source it came from, so it is checkable rather than a confident guess, which matters enormously in a setting where a wrong answer is not just embarrassing but unsafe.

The reason this matters more than it looks is drift. A general model, asked the same clinical-admin question twice, will happily give two confident and slightly different answers, and in a healthcare setting that is not a quirk, it is a liability.

Grounding every answer in the clinic's own written procedure kills it. The system is reading your rules and quoting them with the source, and when a protocol changes, a new intake requirement, a revised rebook cadence, you change it in one place and every answer changes with it.

It is also where the intake checklist becomes enforceable rather than advisory, because the refusal fence in the next step reads its gates from exactly here. The mind is not a chatbot for patients.

It is the single source of truth the whole clinic, and the AI working inside it, reads from.

Step 7. The front-desk employee: filing, chasing and rebooking that actually happens

Now the part people picture when they hear "AI". An employee, not a chatbot, and the distinction is the whole thing.

A chatbot waits to be asked a question. An employee wakes on a timer, reads the live clinic, does its round, files the referrals, drafts the follow-ups, flags who is lapsing, and hands the humans only what needs a human, whether or not anybody prompted it.

I have written the longer version in AI employees, not chatbots.

In a clinic it is a small crew of them, each with one job, all reading from the clinic's mind. A front-desk assistant that reads a photographed referral letter or scan report, understands it, files it to the right patient record and starts the intake checklist.

A refusal fence that will not let a post-op patient be booked in until every required intake item is present, so nobody complex ever slips onto the treatment table before the clinic knows what it is dealing with. A follow-up drafter that watches the clocks and drafts the rebook message, the package-expiry reminder and the discharge check-in for a person to approve.

An invoice filler that drafts the bill for a delivered session straight from the record. A talking agent that answers a patient's "when is my next appointment" or "how many sessions do I have left" in their own language, without pulling a physio off a patient to reach for the phone.

There is a quieter agent in this crew that clinic owners never ask for and always end up valuing most: a critic. Before any piece of work reaches a human, a second agent grades it against the procedure.

Was the referral filed to the right patient. Does the intake checklist actually have all six items or did one slip through.

Is the follow-up message going to the right person at the right point in their plan. Anything that fails goes back to be redone before you ever see it.

This is the line between AI you can run in a clinic and AI you can trust in a clinic, where a mistake is not a typo, it is a patient booked in wrong or a message sent to the wrong person.

An employee that can act is useful and it is also dangerous, especially near patient records and patient-facing messages, so this step ships with four safety rules and they are not optional. They are the reason you can hand a system this much and sleep.

• It refuses out-of-procedure work. No post-op patient gets booked in until the intake checklist is complete, and any referral or request that does not match the written procedure is declined and escalated rather than improvised. The refusal fence is a feature, not a failure, and in a clinic it is the feature that keeps a complex patient off the table until the clinic knows what it is treating.
• Every file operation stays inside set boundaries, and deletes go to a recycle bin. It cannot reach outside the patient records and folders it was given, and nothing it removes is ever gone. A referral filed to the wrong record is always recoverable, which in a healthcare setting is not a nicety, it is the baseline.
• Every patient-facing message waits for a human yes. It drafts the rebook prompt, the package reminder, the discharge check-in, the invoice, then stops. A person reads it and presses send. Anything that touches a patient, their care, their money or the clinic's name gets a human on it before it leaves the building.
• Ambiguous figures and dates come back empty, never guessed. If it is not sure which patient a referral belongs to, how many sessions are left in a package, or when a review is due, it says so and asks. A blank is safe. A confident wrong number about a patient's care or their remaining sessions is how a system does real damage.

Step 8. Hand it to the team

The build is running. Now it stops being the owner's private tool and becomes how the clinic works, and the shift is subtle but total: people move from performing the admin to approving it.

The system does the first pass of everything, the filing, the follow-up drafts, the rebook prompts, the invoices, and a person says yes, tweaks a message, or sends it back. Same front desk, same physiotherapists, far more patients held in proper care, and your qualified people are left doing the assessment, the hands-on treatment and the clinical judgement that was always the actual job.

That handover is made of a few specific things. A daily status drafter that writes the follow-up a patient is owed, ready for a person to approve.

An activity wall so the owner can see what the system and the team did without interrupting a treatment. A change watcher that notices when a patient stalls, when a package is about to expire with sessions inside, or when a referral has sat untriaged over a weekend.

And the piece everyone feels first: the morning brief.

At eight in the morning, before anyone opens the diary, one message lands in order of what matters: here are the two patients about to drop out of care, one a post-op knee nine days without a rebook, here are the three packages expiring this month with sessions still unused and the reminders already drafted, here is the delivered work waiting to be invoiced, here is the new referral stuck two days without triage. The front desk and the physios walk in already knowing the day instead of spending the first hour discovering it, and no finished patient lapses simply because the morning got busy.

A word on the human side, because it decides whether any of this sticks. The word removal frightens a team, and if the front desk and the physios think it means removing them, they will quietly starve the system of the knowledge it needs, they will not tell it about the exceptions, they will not correct its drafts.

It does not mean that. It means removing the data entry, the phone tag, the frightened record hunts, the follow-up guilt, and leaving people with the patients and the care that were always the point.

Say it out loud, early and often, and mean it. A team that believes the system is on their side will feed it every rule it needs.

A team that fears it will fight it, and win.

Step 9. Removal is the whole point

Here is the step everyone gets wrong, and it is the reason the framework exists. Step 9 is not training.

It is removal. The goal was never to teach the owner to use a clever new tool between patients.

It was to take the operating-system job out of the owner's head, so the clinic stops running at the speed of one person's attention, so the follow-up happens whether or not the owner remembered, so a full treatment day no longer means a leaky front desk.

You know you have reached it by a specific test: the owner can take a genuine week off, on a course, on holiday, unwell, and nothing leaks. Referrals still get triaged and filed.

Finished patients still get their check-in. Packages still get watched and nobody's sessions expire unused.

Rebook windows still get a message inside them. Sessions still get invoiced.

Patients still get answered in their own language, the same day. Nothing waits for one person to come back and look, because that person is no longer the thing the routing runs through.

They approve the exceptions from their phone, or they do not, and the clinic holds either way.

And this is where the baseline earns its keep. Every Monday a proof report goes to the owner and says, in the numbers they agreed at the start, what the system did this week.

Referrals filed in minutes instead of the gaps between patients. Every finished patient contacted for a rebook, so nobody dropped off unnoticed.

Zero packages expired with sessions still inside while nobody was looking. It is not a dashboard somebody has to remember to go and check.

It is the system reporting to the owner, unprompted, on whether it is still earning its keep.

And then the question that is really the point: what does the owner do with the attention they just got back. In every clinic the honest answer is the same, and it is why this belongs on a growth marketing site.

They go and do the work only the owner can do. Treating the patients they trained years to treat, without their head half in the diary.

Building the referrer relationships that fill the diary in the first place. Opening the second room, hiring the next physio, deciding whether there is a second clinic.

Removal is not the end of the owner's involvement. It is the first time the owner gets to actually be the owner of the clinic rather than its busiest and most expensive receptionist.

What it costs to run

This is the question every owner asks within about ninety seconds, and the honest answer surprises people in the right direction. It is far less than they expect, and far more transparent, because the running cost is metered and shown like a utility bill rather than buried inside a flat monthly fee that only ever goes up.

For a working system on a real clinic's volume, the AI usage runs on the order of a few dollars a day, and the hosting and database are a few tens of dollars a month, so the whole thing comes to roughly a hundred dollars a month to run. Your volume will differ, which is exactly why the meter is built in from day one and shown to the owner on day one.

No hidden meter, ever, because a clinic owner who cannot see what a thing costs will not trust it near their patients or their money.

Now put that next to what it replaces, with round numbers you can redo with your own. Say the owner and the front desk lose two and a half hours a day between them to chasing referrals, rebooking by phone tag, filing records by hand and remembering who needs a follow-up, across twenty two working days.

That is fifty five hours a month. Value that time modestly, and it is well over two thousand dollars a month of attention, most of it the owner's, against roughly a hundred dollars to run the system that gives most of it back.

And that ignores the two biggest wins, because neither shows up as time saved: the finished patients who came back because they finally got a call, and the packages that got used because somebody watched the clock. Those show up as revenue that did not leave, and in most clinics they dwarf the hours.

Where to start on Monday

Nobody builds all nine steps at once, and you should not try. This is a map, not a project plan, and the map is useful the moment you have it, because it tells you where you are and what the next single step is.

So do not start with the screen that impressed you most. Start with the step the audit said is bleeding worst.

If finished patients keep lapsing and never coming back, start with the follow-up clocks and the follow-up drafter, and you will feel the return within a month as old patients rebook. If untriaged referrals and record chasing are eating your front desk, start with the front-desk assistant and the intake refusal fence.

If you keep delivering sessions and never invoicing them, the money-on-the-floor wall and the invoice filler pay for the whole build almost immediately. The full logic for choosing the first thing is in what to automate first in a service business, and the honest way to measure whether it worked is in true ROI versus reported ROAS.

One thing at a time. Diagnose, build the one piece that hurts most, put a human gate on anything irreversible or patient-facing, then the next piece, which will be cheaper than the first because it sits on the same spine.

That is the whole method. This same framework runs in an accounting firm, a law practice, a business-setup desk and a construction consultancy, the labels on the screens change and the nine jobs do not, and the full walk-through of the general version is in the 9-step framework for deploying agentic AI.

The clinics that end up with a real system instead of a browser full of half-used tools all started the same way: one step too small to fail, aimed at a number they had already agreed was bleeding.