Personal Workflow ProductBehavior TrackingPrivate Data Control

Private AI Gym Coach

A side project that still reflects product thinking and workflow design.

Developer / Designer2026/01 - 2026/04Personal Project

Private AI Gym Coach · Product Flow

Private Data OwnershipCloudflare Workers + D1Selective AI Support
  1. 01

    Workout Log

    Low-Friction Daily UI

  2. 02

    AI Mapping

    Normalize Free-Text Input

  3. User Owned03

    Private D1

    Cloudflare Workers & D1

  4. 04

    Volume Trend

    Weekly Load & Progression

  5. 05

    AI Coach

    Context-Aware Prompts

  6. 06

    Next Plan

    Repeatable Personal Flow

Uses AI selectively where it removes logging friction while keeping core UX fast, clean, and user-owned.

React + TS · Full Data Control
Workflow overview. Some names and data are anonymized or recreated for confidentiality.

Before

Existing gym apps felt generic, cluttered, and too controlling over personal data and daily tracking habits.

After

A private training workflow was built around custom tracking, lightweight AI support, and a UI designed for repeated daily use.

Context & Background

  • This started from a personal frustration with existing fitness apps rather than a client requirement.
  • The interesting part was still workflow design: making a product worth returning to every day.

Constraints

  • The product needed to stay simple, personal, and low-friction.
  • AI support had to be helpful without turning the app into a gimmick.

Problem Solved

  • Commercial fitness apps often introduced more friction than value for daily tracking.
  • Personal data ownership and repeated use experience mattered more than feature volume.

My Role

  • Product design, interface design, and implementation were handled end-to-end.
  • AI was used selectively, with straightforward product design carrying the rest.

System Design

  • Combined workout tracking, personal data storage, lightweight AI mapping, and coaching suggestions.
  • Kept the workflow focused on repeated daily use instead of feature sprawl.

Implementation Highlights

  • Built a custom product around a real repeated-use workflow.
  • Used AI only where it reduced friction, such as exercise mapping and coaching prompts.
  • Designed the UI to feel personal and distinct rather than template-driven.

Outcome & Impact

  • Created a personal product that better matched my own behavior and preferences.
  • Served as a smaller-scale example of turning workflow friction into product design.

Tech Stack

ReactTypeScriptCloudflare WorkersCloudflare D1OpenAI API