Launching today

Fit Receipt
A private fitting agent that knows when to call JEV
11 followers
A private fitting agent that knows when to call JEV
11 followers
Try on lingerie virtually; every pick has an AI judgment receipt. Fit Receipt is a reference implementation. The agent handles chat; code protects facts and pricing. JEV (TypeSafe) is a judgment model that scores trade-offs with calibrated confidence in ~300ms—not essays. Low-confidence results stay unresolved; judgments are recorded; works without JEV. Photos stay in your browser; no server photo storage. JEV sees typed fields, never the photo. Rate-limited API. Vercel + open source.












I would just love to try this out.
For me, it is always a juggle to find a bra that really fit well. To be able to try it virtually and in your own private space would for sure support.
Do you also get like help and support from an AI agent to find the best fit?
Second Eyes
@hanna_curman1 Yes — that's exactly it, Hanna!
The agent asks about your needs (back closure? nude shade? no visible lines?), Jev scores every piece with confidence, and anything unsure goes to an explicit reason instead of a guess. You can ask the agent to do the virtual try and put the result in the fitting room for your review.
All in your browser, nothing uploaded. Try the live demo above and tell me where it gets you stuck? 👙
Second Eyes
Hi Product Hunt! I'm Crystal, founder of NUDE, a Taiwanese intimate apparel brand.
To buy or not to buy — that is always the question for a shopaholic. 🛍️
After years in our fitting rooms, I've learned there is rarely a 100% "right" bra. Preference and habit matter, but so do things no algorithm sees: the mood in the room, the light, how the person helping you makes you feel. A good store follows a quiet order: you see the piece, we learn how you'll wear it, you try it on, and only then do we talk about the cost.
Fit Receipt brings that order online, along with the part of good service that's about honesty.
The agent understands the person. JEV, a judgment model, weighs the trade-offs between styles. Code protects the facts: product data, prices, promotions. The person owns the decision.
Every judgment leaves a receipt, uncertainty included. When the top option sits at 69%, it still says "trade-off still to confirm."
We don't cure choice paralysis. We show you when it's justified.
I'd love to know: where should AI explain its reasoning and where should it make room for irrational human shopping behavior?
Second Eyes
We're live! 🔥 Day 1 update: the Jev integration above is running in production right now — every judgment on the page carries a real receipt. I'll be in the comments all day: tell me where your agent gets unsure, and I'll tell you where ours hands off.