Building for Clinicians Is Not Building for Students

When we started building PhysioAI for practising physiotherapists, we assumed a clinician was simply a student with a licence. We were wrong in five ways, and none of them was about how much the user knows. These are the notes on what changed.

Key takeaways

• A clinician is not an advanced student. Same AI and same knowledge, but a completely different job.
• Students have time and want the why first. Clinicians have a patient waiting and want the what first.
• A wrong answer costs a student a mark. It can cost a clinician a patient, so safety checks come before any workup.
• Clinicians trust answers they can verify, and they need output they can hand to a patient.
• AI can lead a student. For a clinician, it has to follow their call.

The assumption we started with

We thought a clinician was a student with a licence. Same AI, same knowledge, just harder questions and a few extra features.

That turned out to be the wrong model. Each of the five differences we found changed what we built, and none of them was about knowledge.

Same question, different person

A student and a clinician can type the same question and need very different things back.

Five shifts

1. Depth vs speed

A student has time and wants the why before the what. A clinician has a patient waiting and wants the what, then the why.

Synergist, the PhysioAI tutor, breaks its answers into concept, practical, exam focus and safety. That is the right order for learning. For a clinician, it is backwards. The knowledge is the same; the order is different.

2. Teach vs protect

A student needs to learn red flags. A clinician cannot afford to miss one.

A wrong answer costs a student a mark. It can cost a clinician a patient. So Case Workup, the PhysioAI tool for practising physiotherapists, screens for red flags before it starts any workup.

3. Clear vs checkable

A student trusts an answer that makes sense. A clinician trusts an answer they can verify.

Professionals have been burned by confident sources before. So every point in a Case Workup differential shows where it came from. If clinicians cannot check it, they will not use it.

4. Learn it vs hand it over

For a student, the answer is the product. For a clinician, the answer is the start of the work.

A clinician's day ends with something a patient takes home. So Case Workup ends in patient-ready documents, such as exercise plans, not just a reply.

5. Lead vs follow

For a student, AI can lead the way. For a clinician, AI has to follow their call.

A professional does not want to be taught. They want to be backed up. Case Workup supports clinical judgement and never replaces it. The clinician has the last word.

What we would tell any founder building for professionals

1. Define users by their worst moment, not their title. A student's worst moment is an exam. A clinician's is a patient getting worse.
2. Price the wrong answer. Work out what a mistake costs each user, then design for that cost.
3. One engine, different defaults. Change the order, the safety checks and the output. Not the knowledge.
4. Let the expert overrule you. If your product cannot be disagreed with, professionals will not trust it.

What to take away

A professional is not an advanced beginner.

If you build for learners, optimise for understanding. Time is on your side, so spend it on the why.

If you build for professionals, optimise for the next decision. Their time is the scarcest thing in the room.

Same AI. A completely different job. We are still learning this ourselves, and we will keep sharing what the build teaches us.

Case Workup for practising physiotherapists

Case Workup is where these lessons show up in the product. You enter a case once; it checks for red flags, backs your decision with research and an optional second opinion, then drafts notes, letters, exercise sheets and invoices on your letterhead. You stay the clinician. Read more about Case Workup.