Back

Jul 2023 – Sep 2023

Gymee — Unified Fitness Intelligence Platform

Gymee is an all-in-one fitness and nutrition ecosystem that consolidates workout logging, professional coaching, and scientific nutritional tracking into a single, high-performance web application. Built with Django 6.0, it serves both individual athletes and professional personal trainers.

Gymee solves the app fragmentation problem in fitness by providing a centralized hub for exercise data, nutritional guardrails, and coach-client relationships.

Gymee — Unified Fitness Intelligence Platform screenshot

Project Details

Technologies

Python 3.12 Django 5.0 SQLite PostgreSQL Django Unfold Chart.js Pillow

Key Metrics

1,746

Exercise library

15+

Muscle sub-groups

Per set

Real-time PR detection

Goal-based

Adaptive macros

The Problem

Most fitness enthusiasts juggle multiple tools: a calorie tracker, a workout logger, an exercise-form library, and WhatsApp/Email for coach PDFs. This fragmentation leads to:

  • Data silos: Gym progress is disconnected from nutrition targets.
  • User friction: Beginners struggle to find the right exercises and understand muscle targeting.
  • Coach inefficiency: Trainers lack a real-time dashboard to monitor adherence and performance.

The Solution

Gymee provides a unified Fitness Intelligence pipeline:

  • Integrated calculator: Users set their biological metrics once; the system generates adaptive daily guardrails (macro targets).
  • Active session logging: Mobile-responsive workout logging with real-time “previous best” context to motivate progression.
  • Coach-client workflows: Coaches can remotely assign and modify workout plans, keeping training consistent and measurable.

Architecture & Implementation

1) Automated exercise ingestion

To populate a 1,700+ exercise library, I built a robust ingest_exercises pipeline that:

  • Syncs with open-source exercise datasets via JSON APIs.
  • Atomically downloads and maps exercise media to local storage using Pillow.
  • Normalizes unstructured inputs into a strict MuscleGroup hierarchy.

2) Scientific nutritional engine

The nutrition module implements the Mifflin–St Jeor equation:

  • BMR: gender, weight, height, and age.
  • TDEE: activity multipliers from 1.2× to 1.9×.
  • Macros: goal-based protein/fat/carb splits (e.g., higher protein targets for weight loss).

3) Real-time progress tracking

The workouts module uses a set-level SetLog system to monitor every recorded set. When a user exceeds their previous best for weight or reps, the system flags a New PR and visualizes it on the dashboard using Chart.js.

Results & Impact

  • Centralized data: Reduced reliance on external calorie trackers and PDF-based workout plans.
  • Guest-to-member conversion: Session-based storage lets guests try the macro calculator and exercise search before account creation.
  • Scalable content: Automated ingestion enables library growth without manual data entry.

Next steps

  • Wearable integration: Sync heart rate and calorie burn from Apple Health / Google Fit.
  • AI form analysis: Browser-based computer vision to detect posture issues during lifts.
  • Social features: Community leaderboards for key lifts (e.g., deadlift PRs).