How it works

From one turn to a wardrobe that knows its evidence.

A guided capture is designed to become a versioned set of measurement estimates, then a garment template, a digital closet, an on-you preview, and size guidance where the evidence supports it. Production phone-scan processing remains gated until a live provider is configured; see service status.

The whole app journey

Six stages. One evidence trail.

Choose a stage to inspect the app screen. The phone keeps the same body record in view while the job changes around it.

Body captureGuided

Turn slowly

In frameVoice and haptics guide the pace

Concept illustration. All values, garments, and results are fictional.

From the fitting room to the phone.

Built like a record

Nothing important arrives without its context.

Every available value can carry how it was derived, how strong the evidence is, which scan produced it, and when that scan happened. Later features read that record instead of rewriting it.

A guided phone scan

The app guides your distance, position, clothing, and light. If the capture conditions are not supported, it names the issue and offers one next action. A scan can be retried without creating a misleading result. Live production processing requires a contracted scan provider; that provider is not configured today.

Raw capture photos are deleted within 24 hours after successful processing and immediately after failed or abandoned processing.

A version, not an overwrite

Each completed scan creates a dated version. A later scan does not erase the earlier estimate. You can see which version is current and remove a scan when you choose.

Estimated: Calculated from the captured geometry and provider method.
Observed: Supported more directly by visible landmarks in the capture.
Inferred: Derived with more assumptions or limited source evidence.

Two ways to use the result

Measurement spec sheets

Create a dated export for a garment maker or your own records. The export includes units, definitions, source type, confidence, intervals where available, and the scan and template versions used.

Per-brand size guidance

Where reviewed garment data is available, Alex Folzi compares the relevant measurement zones, garment construction, and your fit preference. The result shows the best-supported size, adjacent-size considerations, missing data, and why confidence may be limited. When the evidence is too weak, the product does not suggest a size.

What stays separate

Room is reduced to private deliberate sharing. Discovery, follows, comments, reactions, and public profiles are unavailable.

Optional model training, brand or merchant sharing, and social sharing each require a separate choice.