Case Study

Designing Human-Centered
Automation for Fitness Coaches

Balancing AI and human expertise in training plan creation

Role

UX Research · Interaction Design
Prototyping · Frontend

Duration

8 months

Tools

Figma · React · TypeScript

Impact

Reduced planning time &
improved workflow efficiency

Problem

The challenge

Fitness coaches dedicate extensive weekly hours to creating individualized training plans. The work is repetitive, time-consuming, and demands frequent performance data review.

Existing tools either fully automate decisions — removing human control — or require entirely manual input. Neither approach respects the coach's expertise.

Central Question

"How to support users to reduce repetitive tasks while preserving their control and expertise?"

Research

Users & Context

Target Users

Fitness trainers (powerlifting / bodybuilding)
Part-time professionals
Managing multiple clients

Key Insights

High motivation, value personal coaching
Planning consumes most of their time
Open to technology but cautious about automation

Research Methodology

Pre-study surveyPrototype feedback sessions6-week in-product studyBehavioral trackingQualitative feedback

26 trainers · mixed experience · managing <10 clients each

Discovery

Problems identified

Repetitive manual workflows

Context switching when reviewing trainee performance

Trust concerns regarding automation

Design

Design principles

Augment, don't replace

Assist rather than replace expert judgment.

Transparency builds trust

Explain how suggestions are generated.

Maintain user control

Enable review, modification, and override at every step.

Reduce effort, not quality

Speed up the workflow without compromising accuracy.

Solution

Three levels of automation

Coaches could choose their preferred level — from full manual control to AI-generated suggestions — at any point in their workflow.

Manual

Baseline

Full coach control. No AI assistance — every value is entered by hand.

Assisted

Recommended

AI pre-fills suggested increments. The coach reviews and applies changes.

Automated

Full automation

AI generates the full plan. Coach accepts, adjusts, or switches to manual mode.

From Lo-Fi Prototype to Implementation

The modal interface evolved across three design iterations based on user feedback from the 6-week study.

Lo-Fi Prototype

Light-mode wireframe exploring the automated suggestion layout with circular progress stats.

Mid-Fi Iteration

Dark UI refine: previous-week context columns added to help coaches judge suggestions.

Final Implementation

Production-ready interface with live trainee data, trend indicators, and one-click accept.

Findings

Key learnings

"Trust is the core UX challenge in AI"

Users reject loss of control, not automation itself.

"Efficiency ≠ fewer actions"

Users remained active but worked faster through improved information design.

"Human-centered automation works"

Optimal results: human decides, system suggests.

"Reduce effort, not quality"

Speed shouldn't compromise personalization or accuracy.

Explore more

More Case Studies

Back to Case Studies

Designing Human-Centered Automation for Fitness Coaches