Skip to project brief
ZENAI← Selected workBUILD / 03
03 / The physio content studio

A small team.
A bigger imagination.

Four AI roles. One business owner. Step inside a creative studio where research becomes a strategy, a weekly content plan and a conversation about what worked.

A content operation built around a physio business and its owner.
Prototype · fictional data
Text walkthrough · no audio
A warm illustrated creative studio with research monitors, campaign boards and an owner’s review table.
IN THE STUDIO / 01Research notes
AI / Alex · Research agent

Find the constraint behind the run

“Across 24 runner check-ins, hamstring tightness after long runs came up most, 9 times. That’s a listening signal, not a diagnosis. Sam should sanity-check it.”

Full reasoning

I reviewed 24 runner check-ins. The most repeated self-described constraint is hamstring tightness after longer runs, mentioned in nine responses. Ankle pain on uneven ground appears in seven, and running-posture questions in five. That is a listening signal, not a prevalence study or diagnosis. Before we turn it into content, Sam should compare the themes with recent enquiries and confirm which questions the clinic can answer responsibly.

INSPECTING / Research / AI

Monitor audience questions and market themes. Keep a source trail and separate observations from assumptions.

Proposed tools Permitted research sources · source library · language model

Introduce the clinic to local runners. Step 1 of 5. Alex · Research agent: Find the constraint behind the run. I reviewed 24 runner check-ins. The most repeated self-described constraint is hamstring tightness after longer runs, mentioned in nine responses. Ankle pain on uneven ground appears in seven, and running-posture questions in five. That is a listening signal, not a prevalence study or diagnosis. Before we turn it into content, Sam should compare the themes with recent enquiries and confirm which questions the clinic can answer responsibly.

Try the system

A working slice
you can touch.

Small interactions make the operating model visible: explicit rules are calculated, AI drafts stay inspectable and people keep the final say.

The weekly plan

Three posts. Three clear jobs.

Learn which running constraint the clinic can explain responsibly, then turn that evidence into useful conversations.

Monday listen · Wednesday unpack context · Friday invite a safe question
How should the clinic sound?
20-second question-led runner reel

The job of this post

Surface the constraints runners actually want help understanding, without assuming a diagnosis or a single cause.

Movement studioPhysio clinic · Draft post
MON / 20-second question-led runner reelWhat gets in the way of a good run?

Runners, what gets in the way of a good run? We’re listening for the questions behind hamstring tightness, ankle pain, running posture and the other things that interrupt training. Share a general question for the team; please keep personal health details out of the comments.

Draft caption. The production brief and checks below travel with this post.

Creative direction
Open on a runner pinning three cards to the studio wall: hamstring tightness, ankle pain and running posture. Ask: ‘What interrupts your run?’ End with a general-question prompt, not a treatment tip.
Call to action
Share one general running question; keep personal health details out of the comments.
Owner checks before release
Confirm the constraint wording with a clinician; approve staff and runner footage; set the reply boundary and escalation route.
0 of 3 posts reviewed. Changing tone resets the reviews.
Sam’s correctionsWhat the owner sent back
Human / clinic owner

Approval stays locked until you’ve resolved: 3 post reviews, a named reply owner, 4 of Sam’s corrections.

Awaiting owner approval.
Behind the environment

Tools are chosen
around the work.

A proposed stack, not a deployed client system.

AIResearch

Permitted research sources · source library · language model

AIStrategy

Business brief · agent orchestration · shared memory

AIWeekly plan

Content database · language model · editorial workflow

AIReporting

Social analytics import · metric calculations · report drafting

HumanOwner review

Approval queue · approved-content scheduler · activity log