Led the end-to-end redesign of a mental-training platform — from fragmented features to a role-based product architecture.
Reframing the product
Every new feature increased the decisions users faced, multiplied role-specific exceptions and created more design and engineering divergence.
The risk was not visual inconsistency; it was a product that became harder to adopt and more expensive to evolve.
Business stakes
Growth meant more athletes, more staff roles and more content. Without a clearer product architecture, every new capability increased onboarding friction and implementation cost.
New users had to interpret the product before receiving value.
Role exceptions multiplied design and engineering effort.
Features had no stable logic to attach to.
The diagnosis
The redesign had to reduce decisions without reducing agency.
Cognitive overload
Blurred mental models
Inconsistent interactions
A system that could not scale
The answer was not a thinner feature set. It was a product model organised around intent, role and momentum.
Designing the system
They gave Product, Design and Engineering a common test for every flow.
The next meaningful action wins.
Reach value before asking for commitment.
Organise around why users arrive.
Reuse logic before reusing pixels.
Mapping capabilities by role exposed where the experience should be shared, where it needed to diverge and which responsibilities were creating unnecessary overlap.
Organise, define and supervise
Manage, monitor and guide
Prepare, practise and compete
The first bet
We shortened onboarding and delayed personalisation until after the first meaningful session. The trade-off was deliberate: less tailoring upfront, more momentum when motivation was highest.
Personalise first
Reach value sooner
Personalise progressively
AI in the process
The product architecture and four principles were defined without AI. Once stable, they became context for flow exploration, edge-case expansion and implementation-ready prototypes.
AI accelerated alternatives. It did not decide what the product should be.
Building the experience
Daily actions and long-term progress became distinct but connected. Navigation communicated what came next, while progressive disclosure kept exploratory routes available without competing with the primary task.
Early iterations were too restrictive for exploratory users. I reintroduced progressive disclosure: the primary action stayed obvious, but deeper layers remained available when the moment required them.
What changed
Across twelve months, I led two product designers, established the four principles and split the redesign into role architecture, activation and system workstreams.
Early iterations over-corrected. I restored progressive disclosure so the primary action remained obvious without trapping exploratory users in a narrow path.
Early validation
The redesign reduced ambiguity in walkthroughs. The next responsibility was proving that clarity changed activation and return behaviour.
A product becomes scalable when its logic is clear enough to survive the next feature.