Valen: Fitness Application  valenapp.tech ↗

Valen is a fitness training intelligence app I built solo end to end, product, engineering, and design, using an AI native workflow with Cursor, Claude Code, and Figma. Built in React Native with Firebase and JavaScript.

React Native Firebase JavaScript Figma Cursor Claude Code Antigravity

The Real Problem

Most fitness apps are glorified spreadsheets. They log what you did but give nothing back. No adaptation, no pattern recognition, no intelligence.

Approach

Problem Scoped

Identified the core lack of adaptation in fitness apps and defined an intelligence-first experience.

System Built

Developed a cross platform React Native app backed by Firebase using an AI native workflow.

Shipped

Released into closed testing with a functioning core loop, cloud backup, and subscription entitlements.

What I Actually Built

I designed a fatigue heat map on the homepage so you can see muscle readiness visually instead of reading tables. For logging, I built a Quick Log mode that records a session in three steps to reduce friction, alongside full logging when precision matters. The system uses a compressed four step onboarding; muscle readiness, training intent, session confirmation, and start; and features a weekly planning system connecting intent to execution. It includes cloud backup and restore for training history, and manages subscription entitlement handling across iOS and Android edge cases.

Key Decisions

Cold Start

Problem: Standard 20+ question setups cause early drop offs.

What I did: Compressed onboarding to 4 steps instead of 20 plus.

Outcome: Delivered immediate value with minimal friction.

Logging Friction

Problem: Detailed logging creates too much mid session friction.

What I did: Built Quick Log for busy sessions, full logging when precision matters.

Outcome: Flexible logging that users actually stick to.

Information Overload

Problem: Dense data tables overwhelm users looking for quick insights.

What I did: Made the homepage visual first with a heat map, data sits behind it.

Outcome: Muscle readiness is instantly readable at a glance.

Feature Scope

Problem: Over scoping features risked delaying the initial release.

What I did: Aggressively deferred anything that would delay the core experience.

Outcome: Shipped a functional, stable core loop in just 4 months.

Real Learnings

Users trust the system more when it explains its reasoning; showing the "why" increased adherence. Biometric variance is high early on, which required cold start logic before the engine had enough history to be accurate. From an engineering perspective, subscription entitlement handling across iOS and Android edge cases was harder than expected. Ultimately, product ownership at this scale requires genuine technical curiosity, not just enough fluency to hand things off.

Currently in closed testing. Live at valenapp.tech.