A workout planner and training log. You tell it your goal, experience level, how many days a week you can train, and what equipment you have; it builds a program. Then you log your sets against that program and it keeps track of your history, PRs, and progress per exercise.
Because I've been lifting weights since 12 years old, and competed in olympic weightlifting in college, Adaptix was a natural project idea for me given my background in fitness and computer engineering. I hope people can take great benefit from using it and improve their mental and physical health.
Built with React + TypeScript on the frontend, FastAPI + Python on the backend, and PostgreSQL for storage.
- Builds a program. Picks a split (full body, upper/lower, etc.) based on your days per week and experience, then fills each day with exercises, sets, rep ranges, and rest times that fit your goal and equipment. This is rule-based, not ML.
- Tracks workouts. Start a day from a saved program and log weight × reps for each set. It shows what you did last time for each exercise, and Use Previous fills it in for you.
- Finds PRs. Personal records are computed from your history, with the previous best and estimated 1RM next to each one.
- Charts progress. Each exercise has its own history page with a chart you can switch between top weight and estimated 1RM.
- Handles lb and kg. Pick a unit and everything displays in it. More on how that works below.
| Workout builder | Tracker | Personal records | Exercise progress |
|---|---|---|---|
![]() |
![]() |
![]() |
More screenshots are in screenshots/.
Weights are stored in the unit they were logged in. Each set log keeps its original unit, and conversion only happens in the frontend (frontend/src/utils/weight.ts). Switching from lb to kg never rewrites the database, and a history with mixed units still compares correctly when finding PRs.
History outlives programs. Deleting a saved program doesn't wipe the sessions you logged with it. Session data keeps enough exercise info on its own to stay readable.
Estimated 1RM uses the Epley formula: weight × (1 + reps / 30). A single rep counts as its own 1RM.
You'll need Python 3.11+, Node.js, and PostgreSQL.
1. Create the databases (one for dev, one for tests):
CREATE DATABASE personalized_workout_app;
CREATE DATABASE personalized_workout_test;2. Add a .env file in the project root:
DB_HOST=localhost
DB_PORT=5432
DB_NAME=personalized_workout_app
DB_USER=your_postgres_user
DB_PASSWORD=your_postgres_password
TEST_DB_NAME=personalized_workout_test3. Install dependencies and set up the schema:
pip install -r requirements.txt psycopg2-binary
python -m backend.init_db
cd frontend && npm install4. Start both servers (in separate terminals):
# from the project root
python -m uvicorn backend.main:app --reload # http://127.0.0.1:8000 (docs at /docs)
# from frontend/
npm run dev # http://localhost:5173python -m pytest -q # backend, runs against the test database
cd frontend
npm run lint
npm run buildThere are 29 backend tests covering the recommender, the generator, the repositories, and the API. The frontend doesn't have tests yet.
The full interactive reference is at /docs once the backend is running. The main routes:
| Method | Route | |
|---|---|---|
POST |
/api/account/start |
Start with an email; creates the user if it's new |
GET PUT DELETE |
/api/users/{user_id} |
User profile |
POST |
/api/workouts/preview |
Generate a program without saving it |
GET POST |
/api/users/{user_id}/workouts |
List or save programs |
DELETE |
/api/users/{user_id}/workouts/{program_id} |
Delete a program |
GET POST |
/api/users/{user_id}/sessions |
List or start sessions |
POST |
/api/users/{user_id}/sessions/{session_id}/sets |
Log a set |
PUT |
/api/users/{user_id}/sessions/{session_id}/complete |
Finish a session |
backend/
main.py FastAPI app and routes
recommender.py picks the training split
workout_generator.py builds each day's exercises
exercise_library.py exercises grouped by movement pattern and equipment
*_repository.py database access
schema.sql, init_db.py schema and setup
frontend/src/
components/ one component per screen
utils/weight.ts lb/kg conversion
tests/ pytest suite
- Real authentication (right now there are user profiles but no passwords)
- Editing programs and adding custom exercises
- Frontend tests
- Deploying it somewhere
Built by Argenis Y. Vélez Álvarez, Computer Engineering student at the Polytechnic University of Puerto Rico.



