CycleSync
Training timed to her, not to her download date.
Inspiration
Almost all sports science was done on men. Women were left out of exercise studies for decades, so the training programs everyone follows were built on male physiology and handed over unchanged.
On top of that, every workout app starts you at week one, whenever you happened to download it. That means your hardest training week lands at random, and when a weight that felt fine two weeks ago suddenly feels brutal, you assume you're failing.
What it does
CycleSync builds a four-week strength and conditioning plan and places the hard week and the deload week according to where you are in your cycle instead of your download date.
You enter your height, weight, goal, training days, equipment, and when your last period started. It picks the exercises, scales the weights, and tells you what to expect before you feel it.
| Week | Phase | Intensity |
|---|---|---|
| 1 | Period | Moderate |
| 2 | Late follicular | Peak, go for PRs |
| 3 | Ovulation to early luteal | Hard, not maximal |
| 4 | Late luteal | Deload |
After each session you log how it went in three taps. Over a couple of cycles it moves your peak week if your data says the default guess was wrong.
How we built it
- React + Vite on the front end
- Auth0 for sign-in
- Two data files carry all the training logic: 86 exercises across three equipment tiers, and the phase rules, prescriptions and research citations
- Deterministic generator, no API calls, nothing to break during a demo
Challenges we ran into
Auth0 and the domain setup ate more of our day than the actual product did. Between callback URLs, environment variables and DNS propagation, we lost hours to config rather than code, which is a frustrating way to spend a hackathon.
The evidence was contested. The research on cycle-based training is genuinely mixed, and the largest study, Apple's with over 110,000 women, found no difference in how much women actually train across phases. We had to decide whether to overclaim or be honest, and we chose honest. The app cites the papers that disagree with its own premise.
We caught ourselves inventing metrics. Our first session log used a made-up 1 to 5 energy scale, so we replaced it with Borg CR-10 and the Hooper fatigue item, which are actually validated.
The exercise library was gym-only. When we tested the equipment filter, a home or bodyweight setup produced a lopsided plan, so we had to build the library out before shipping the feature.
Accomplishments that we're proud of
- It fully works. Every screen loads, nothing is faked.
- A real exercise library with three equipment tiers, so it isn't only an app for people with a full gym membership.
- It tells the truth about what it does and doesn't know, which almost no health app does.
What we learned
Knowing where your evidence stops is more valuable than sounding certain.
We started out wanting to claim the app makes women stronger faster, and the research wouldn't support it. The honest version, that it makes training make sense so a hard week reads as expected instead of as failure, turned out to be the better product anyway.
What's next for CycleSync
- More cycles of data so the personalization has real history behind it
- Perimenopause support, where the cycle stops being predictable and the need is even bigger
- A deeper hormonal birth control mode that goes beyond the steady four-week block we ship today ```
Log in or sign up for Devpost to join the conversation.