Paizah

Paizah Counter: a laundry counter on WhatsApp, checked by Meta's models. Laundry counters in India run on desktop or mobile billing software: staff count each customer's lot by hand and key it in. Counts go wrong, and a stain found after washing becomes the laundry's fault. At Prezr, a laundry in Thiruvananthapuram, we moved the counter onto WhatsApp: staff send a photo of the lot and say what is in it, and the quote is made in Zoho Books. Paizah Counter is the step that makes this a product any laundry can adopt, with the checks a counter needs. At the counter, it reads the photo and the staff's words separately with Muse Spark, compares them garment by garment, and prices the lot from the laundry's own price list in code. When the two agree the quote is ready; when they disagree, staff are asked first. On the floor, before washing, Muse Spark records stains and tears and SAM 3.1 circles them on a counted photo that goes to the customer. At handover, the delivery photo is counted against what was billed at intake, so a missing piece is caught before the bag leaves. At Prezr the counter count is what everything after it runs on: the Zoho Books quote, the tag on every piece, the floor stages, the Zoho invoice sent on WhatsApp, and delivery. Those systems existed before this hackathon and are not part of the entry. Tested on 111 real Prezr counter quotes scored against final invoices: when words and photo agree (three lots in four) the count is right 97.5% of the time, and the check caught 19 of 21 staff counting errors. About 3 US cents a quote.