Case study — Product & Tech
Sweat
Alliance
Building the brain behind the sessions.
Sweat Alliance's workshop curriculum existed in a trainer's head: a constellation of movement corrections, coaching cues, and session structures with no shared language and no way to scale.
I productized that knowledge — translating informal expertise into a structured fitness design system that could be taught to new trainers, communicated to clients, and embedded into the business's growth model.
-
Role
Sole designer & system architect
Concept to launch, self-initiated
-
Timeline
6 months design, ~1 year live
Workshops running at ~70% capacity
-
Deliverable
Paid monthly workshop series
+30% repeat attendance
Expertise that couldn't travel.
The head trainer could read a squat pattern in seconds — but that skill lived entirely in muscle memory and intuition. No documentation. No taxonomy. No repeatable framework.
When I audited the existing curriculum I found three recurring friction points:
- No shared vocabulary. Different trainers used different terms for the same fault. Clients received inconsistent corrections across sessions.
- No fault-to-cue mapping. Knowing what was wrong didn't automatically surface the right correction. Each trainer improvised, creating quality variance.
- No onboarding path. New trainers took months to reach the head trainer's assessment quality. Scale was capped by the person, not the system.
Three people. One system.
Research surfaced three distinct user types — each interacting with the system at a different layer of abstraction. The design had to serve all three simultaneously.
-
The Trainer
In-session practitioner
"I need to spot the fault, name it, and cue the fix — in ten seconds, mid-class."
Needs a fast fault-to-cue reference that's pattern-based, not memorized. Precision under pressure.
-
The Client
Active learner
"Why do I keep hearing the same correction? I want to understand what my body is doing wrong."
Needs conceptual framing, not just commands. Understanding leads to buy-in, which leads to behavior change.
-
The Owner
Business operator
"I want new trainers to deliver at my level without me in the room every time."
Needs a scalable knowledge artifact — training material that lives beyond any single session.
7 patterns. 6 fault types. 42 cue trees.
I mapped every movement the curriculum covered to a primary pattern category, then classified observed errors into six fault types across each. Each fault maps to a standardized coaching cue.
Select a movement pattern to expand its fault taxonomy below.
From freeform notes to structured data.
Session notes lived in a mix of voice memos, whiteboard photos, and trainer shorthand. I defined a metadata schema that normalized session observations into queryable, trainable records.
Before — Freeform session log
- Note
- "Jordan's squat is still off — butt wink again, told him to brace harder"
- Date
- Tues (no year)
- Fault type
- Not recorded
- Cue used
- Not recorded
- Outcome
- Not recorded
After — Structured session record
- Pattern
- Squat
- Fault type
- Posterior chain — Lumbar flexion (butt wink)
- Cue applied
- "Chest up, push the floor away"
- Client response
- Partial improvement — requires mobility work
- Follow-up flag
- Hip flexor assessment — next session
A year in. The system held.
The workshop series launched as a paid monthly product — and stayed live for over a year. The design system did what it was supposed to: the knowledge traveled.
-
70%
Workshop capacity
-
30%
Repeat attendance lift
-
12+
Months live & running
- Trainers onboarded faster. New staff reached assessment fluency using the taxonomy as a training scaffold — no shadowing months required.
- Clients progressed more consistently. Shared vocabulary meant corrections carried across sessions and trainers — clients heard the same language, built the same mental model.
- The business scaled past the founder. Revenue from workshops became a predictable line item — not dependent on one person's presence or availability.