A cycle and symptom companion for people with PCOS, whose periods can be weeks late or not come at all.
CycleSync is an iOS concept. PCOS is a hormonal condition that makes periods irregular, late, or absent, and most tracking apps are built around a regular month. I designed this over six weeks, starting with a survey of ten people and ending with a one-page summary they can hand to a doctor.
View prototype
PCOS is a hormonal condition that affects about one in ten people worldwide. It makes periods irregular, late or absent, and up to 70% of people who have it are undiagnosed.
Most cycle trackers are built around a regular 28-day month. The brief I set myself was an app for people whose cycles do not fit that month, that keeps a record a doctor will take seriously.
A tracker counts down to a predicted date. With PCOS the real gap between periods runs from 35 days to more than 90, so the countdown passes zero and turns into a warning.
Keep long cycles visible across the whole year, so a 60-day cycle reads as one bar and not as a missing month.
Take symptoms as a typed sentence, sort them into flow, mood, symptoms and energy, and let the person check the result before it saves.
Put cycle history, recurring symptoms and the last few months of patterns on one page for the doctor.
I ran an open survey with ten people who have PCOS, asking how they track their cycle, what they were told at diagnosis, and how they prepare for an appointment.
Each product on the market does one part of the job. Flo and Clue track cycles, AskPCOS explains the condition, Allara and Pollie connect people with clinicians. None of them produces a record a person can take into the room with a doctor.
| Tool | Good at | Falls short |
|---|---|---|
| Flo and Clue | +Polished, large content libraries | –Calendar assumes a regular month –No visit summary or doctor handoff |
| AskPCOS | +PCOS-specific, evidence-based | –Information only, no tracking –No record of the user’s own pattern over time |
| Allara and Pollie | +Real clinicians, personalised care | –Paid telehealth, so out of reach for most people |
| Generic AI chat | +Fast, conversational | –No citations or reliable symptom record |
Six personas. Maya and Aisha from published research, the other four from the survey, quoting the responses directly.
Maya“The app should fit my cycles, not force them into a standard month.”
Diagnosed four years ago, after three doctors. She uses a mainstream period tracker, and her cycles do not fit its month.
The flow runs from the first open to the doctor's appointment.
The palette is warm on purpose, because this is an app someone opens on a bad day. Luna answers in plain language with a source on every claim, and says so when the evidence is thin.
The in-app companion. Luna answers in plain language, attaches a source to every claim, and says so when the evidence is thin.
Six screens, from first setup to the page a person hands to a doctor.
Each one answers something from the survey: the prediction that was wrong, the note typed into a phone, the question nobody answered, the appointment that ended without a plan.
Home shows today. The cycle day, how the person feels, what they logged, and one note from Luna. There is no countdown, because the app cannot know the date.
Three questions, and a skip for anything that assumes a diagnosis.
Twelve months on one screen. A 60-day cycle is a long bar, and a month with nothing in it stays empty.
Luna answers questions in plain language. Every answer names its source, and when the evidence is thin Luna says so.
The daily log takes a sentence typed the way a person would say it and turns it into flow, mood, symptoms and energy. The person checks the result before it is saved.
The visit summary puts cycle history, recurring symptoms and the patterns of the last few months on one page. It is the record people were already trying to build in their notes app.
The prototype connects the primary flow from onboarding to the visit summary. The remaining states are included for review in Figma.
I started out trying to predict the next period better. The survey changed that.
Nobody asked for a better prediction. They asked to be believed in a room with a doctor, and a clear record is what makes that possible.
Next is to test the year view with people who have PCOS, refine the visit summary with a clinician, and look at whether a doctor consultation can live inside the app rather than as a separate handoff.
A camera-first repair application that helps users repair everyday objects, and defers to a professional when a task is unsafe.
