Morie
Morie is a cycle tracker for people whose cycles do not run on a schedule. It logs the symptoms other apps have no field for, reads lab results against them, and surfaces the patterns between the two.
Symptoms with no field
Cycle tracking apps are built for a cycle that arrives on schedule. They carry fields for flow, mood, and cramps, and they predict forward from an average.
For someone with PCOS or endometriosis, the average is the least useful number available, and the symptoms that matter most have nowhere to go. Brain fog. Energy crashes. Sugar cravings. Hair changes. Skin flare-ups.
The cost is not a worse app. It is sitting in an appointment describing six months of symptoms from memory to someone deciding whether to order a test.
A day’s entry. Condition-specific symptoms rated by severity rather than checked off.
Not a tracker
We surveyed 54 people in January 2026, recruited through PCOS and endometriosis groups on Facebook, Reddit forums, and the reach that followed from those posts. The plan going in was a tracker with fields the other apps did not have.
Most respondents were already tracking. In apps, in notes, in their heads. Collection was not the problem. Nothing they collected added up to anything they could use, and the place it failed was the exam room.
That ended the tracker-first plan. If the product is evidence rather than a log, it needs the two things a log does not have: the clinical numbers a doctor already accepts, and enough context that a person knows what to ask.
Lab results came out of follow-up conversations with respondents. People with these conditions have panels. Testosterone, fasting insulin, thyroid. Those numbers sit in a portal, disconnected from anything the person felt that month.
So Morie is three things in one app, and the reason they are one app is that each is useless to the others at a distance.
The product
A field for every symptom
Severity on a scale rather than a checkbox, plus free text for anything the taxonomy misses. A symptom that only shows up as present or absent cannot show a trend, and trend is the entire point when the thing being described is gradual.
Cycles that vary as the baseline
The standard cycle visualization is a ring, and a ring encodes an assumption: that the period is fixed and the only question is where you are inside it. For an irregular cycle that assumption is the error.
Period history draws each month as its own track, with the period placed where it actually fell and drawn at the width it actually ran. Variation reads as the shape of the data rather than as a deviation from a norm. A missed month is drawn as an outline in position rather than left blank, because for this population a missed cycle is a data point, not a gap in the record.
What the AI reads
The pattern layer works across series rather than within one. A symptom that trails the period by a week is not visible in either series alone, and it is the kind of observation someone would otherwise have to notice themselves across months of entries.
It reads one person’s logs against that person’s own history and their own lab values. Health data is encrypted. Nothing is sold, and there is no cross-user model in the loop.
Observations surface in their own tab rather than as notifications, each one tied to the window of days it was drawn from, so a claim about a pattern can be checked against the entries that produced it.