Movement
PractitionerLedger layerMulti-tool stack

Listening for the Gap Between Sensation and Telemetry

How a movement therapist used language models to map the friction between client sensations and sensor figures.

4 min readWellness & AI editorial

A physical therapist running a boutique rehabilitation practice in Northern Europe found herself overwhelmed by contradictory client updates. Between sessions, clients would send hurried voice notes describing sudden knee stiffness or shoulder fatigue, often directly contradicting the calm recovery figures shown on their movement trackers. By Thursday evenings, her desk was covered in scribbled post-it notes and cross-referenced calendars, leaving her second-guessing whether she was under-loading recovering joints or permitting clients to train straight through genuine tissue distress.

She stopped trying to referee the conflict between subjective complaints and numerical targets. Instead of treating client discomfort as an anomaly to be measured away, she began viewing the gap between lived sensation and mechanical output as the primary diagnostic conversation. The client’s own vocabulary became an active clinical compass.

She spent three weeks transcribing client movement debriefs into an unformatted evening log alongside wearable recovery summaries. Rather than building intricate classification schemas, she used an analytical language model as an impartial conversational mirror twice a week. She asked it simply to contrast what clients reported feeling against their recorded movement volumes, looking exclusively for moments where client descriptions diverged sharply from the objective load recorded by their devices.

Clients stopped arriving at appointments with defensive explanations about missed exercises, instead beginning their sessions by pointing calmly to specific warm-up movements that felt out of sync with their baseline.

Adapt the shape to your own stack

Vendor-neutral steps. Use whichever AI tools you already trust — the shape of the work matters more than the brand.

  1. 1

    Aggregate sensory notes and sensor data in one document

    Maintain a simple ongoing log where subjective post-session comments sit directly beside objective movement metrics.

  2. 2

    Query a language model for perceptual discrepancies

    Prompt an analytical model to scan your weekly entries solely for instances where subjective fatigue and mechanical load diverge.

  3. 3

    Isolate language patterns linked to overreaching

    Review the flagged passages to identify recurring sensory phrases that consistently precede drops in physical tolerance.

  4. 4

    Adjust the training conversation, not just the volume

    Use the discovered discrepancies as discussion prompts during check-ins to calibrate pacing collaboratively with the mover.

Three things to read next.

See all →

Suggested for you

Based on what you've been reading — always learning.

See all →