About Physisync
We build the accuracy layer that stops apparel returns before the parcel leaves the warehouse.
What we do
In Indian direct-to-consumer apparel, a large share of cash-on-delivery orders comes back to the seller as RTO — return to origin. The brand pays freight in both directions, carries the stock lock-up, and learns nothing from it, because a courier manifest has around eleven reason codes and not one of them can say it ran small. A size failure is filed as a refused delivery, so the money gets budgeted against fraud and address hygiene, and the pattern that caused it never changes.
Physisync re-cuts that spend against the brand’s own records. The attribution runs on order history, exchange reasons and size charts a brand already has — no courier integration, no manifest change, no re-tagging of historical orders — and returns a ranked list of styles whose size charts are provably wrong, each with a rupee figure attached.
The correction then lands where the mistake is actually made. Each style earns a learned centimetre offset that the size picker applies before checkout, with optional camera-based body measurement that runs on the shopper’s own device. What sizing cannot catch is caught before dispatch: a WhatsApp confirmation on high-risk orders, and a Hinglish voice call when that goes unanswered. Both fire before pickup, which is where the economics are — a cancellation before dispatch costs nothing, an RTO costs two legs of freight.
How we build
- Nothing new to instrument. Every number we produce comes from data the brand already owns. If a claim needs a new integration to be true, it is not our claim.
- Measurement stays on the device. Camera sizing runs in the browser. Body imagery is not uploaded and not stored.
- The numbers carry their assumptions. Our published figures come from a synthetic corpus with a retained answer key, and the site says so on the page where the figures appear. We would rather be checkable than impressive.
Company details
Get in touch
Brand teams looking at their own RTO line, and anyone who wants the method walked through in detail, can write to hrutik@physisync.co.in. The working demo — attribution, size picker and pre-dispatch confirmation, end to end — is linked from every page of this site.