Ashvara
Case study

Cycle tracking that never leaves the phone

How we built Lunara as a period tracker with no servers at all - every prediction computed on-device, so the most sensitive data a person holds is never collected.

App
Lunara
Our role
Design & build
Platform
iOS
Year
2026
Live
on the App Store
No servers
data never leaves your iPhone
No account
nothing to sign up for

The problem

Cycle and reproductive data is some of the most sensitive information a person holds - and in 2026 it's data people are right to be nervous about handing to a company. The usual answer is a longer privacy policy and a promise to guard the servers well. We think that's the wrong answer. The safest way to protect data isn't to secure where it's stored - it's to never collect it in the first place.

We wanted Lunara to be a genuinely private period tracker: fast, accurate, and calm to use, with a privacy story that isn't a promise but an architecture.

Constraints

We set three rules before designing a screen:

  • No servers, no account. If there's no backend, there's no database to breach, subpoena, or sell. The app should work the moment it opens, with nothing to sign up for.
  • Predictions still have to be good. "Private" can't mean "dumb." The cycle, flow, symptom, and mood predictions had to be computed entirely on-device and still feel accurate.
  • Pay once, if at all. No subscription for something this personal - a free app with a single optional unlock.

The strongest privacy guarantee isn't a policy you have to trust. It's a design where the data was never sent anywhere to begin with.

What we built

Lunara logs a cycle, flow, symptoms, and mood in seconds, and forecasts the next period and fertile window from your own history - all computed locally on the device. There's no login screen because there's no account. There's no "syncing" spinner because there's nothing to sync to. The whole experience is designed to feel quick and reassuring, with a clear on-screen reminder that the data stays on the iPhone.

The predictions run against your logged history using an on-device model of your averages and ranges, so the estimates sharpen as you use it - without a byte of that history leaving the phone.

The tech

A native SwiftUI app with a fully local data model and on-device prediction - no network layer for user data at all. Monetization is a single optional Pro unlock through StoreKit, so the business model never depends on collecting or monetizing the data. The design is warm and low-stress, built to make a sensitive daily check-in feel light rather than clinical.

The outcome

Lunara shipped to the App Store as a period tracker whose privacy is structural, not promised: no servers, no account, predictions computed on-device, and an optional one-time unlock instead of a subscription. It's proof that "private by default" and "genuinely useful" aren't a trade-off - they're a design decision.

Have something to build?

We'd love to hear about it. Tell us what you're working on and we'll take it from there.