Making a patient-facing app feel faster
How measurement-led frontend refactoring improved a critical experience without a risky rewrite.
Roughly 40% faster load time on the targeted experience
The problem
A patient-facing workflow was slow enough to affect usability. The useful goal was not a perfect synthetic score; it was reducing the wait before people could complete an important task.
The investigation
I established a repeatable baseline and traced the critical rendering path before proposing changes.
- Profiled expensive renders and unnecessary work
- Compared production-like devices and realistic data
- Separated network, JavaScript, and rendering costs
- Prioritized delays users could actually perceive
The decision
Focused refactoring had a smaller blast radius and clearer measurement than a rewrite. Changes were released incrementally so each improvement could be verified.
What I learned
Performance is a product feature. A successful optimization starts with a user-visible symptom, changes the constrained path, and keeps regression protection after release.