REPDAILY

The tracking mechanic was technically achievable, but the real brief was user retention.

Most fitness apps hemorrhage users within two weeks because they fail to provide a compelling reason to return.

Additionally, the camera-based calibration had to be flawless—if the initial setup states felt awkward, the product would fail at the exact moment it needed to work.

Before any UI was designed, I mapped the complete product ecosystem across six functional flow maps: onboarding, calibration, core training loop, progression, dashboard states, and CRM lifecycle.

To de-risk the build, a beta committee of personal trainers and early users ran structured feedback loops.

This live testing directly reshaped the calibration flow and dashboard hierarchy.

The product strategy also required a deliberate Free vs. Paid split; rather than crippling the free tier, the Pro version was designed to offer meaningful, structured progression mechanics.

RepDaily was a live stress-test of a modern, lean engineering workflow.

Alongside Figma for system design and token management, I used a cross-LLM prompt development process to iterate UX rationale and generate comprehensive PRD scaffolding.

Crucially, Cursor was deployed to rapidly prototype test sandbox modules.

By exploring live-coded design concepts, edge cases, and UI interaction states directly in code, we bypassed the taxation of traditional wireframing.

This AI stack was not a shortcut; it was a watertight pipeline that allowed a two-person team to execute at the scale of five.