“"The 3D body avatar has 7k vertices, with unlimited measurement possibilities... Tighter clothing works best, just like when you're getting fitted by a tailor." — Yuri Douglas (yuridoug), founder, 2024-08-09 (source: HN Algolia API, hn.algolia.com/api/v1/items/41204737; live HN ...”
You know that feeling when you order clothes online, they arrive, and they still don't fit? Online apparel return rates run 20–30%, and fit failures account for 53% of those returns per industry reports — meaning the sizing problem, not buyer's remorse, drives the majority of returns. Existing solutions either require physical kiosks and scanning hardware, or ask shoppers to self-report measurements they don't know, producing data that's wrong before the first API call.
You take a front-facing photo and a side-facing photo on any smartphone. Both are encoded as base64 and sent to SnapMeasureAI's servers, where a probabilistic 3D reconstruction pipeline combines the depth cues from both angles to build a 7,000-vertex body mesh. The system extracts 100+ measurements across 10,000+ indexed body points and returns results in under 10 seconds. The mesh can also export as a standard OBJ file for downstream use. Per the founder's own submission (retrieved via HN Algolia API), images are processed and discarded — never stored.
If you're a backend engineer or product lead at an e-commerce apparel brand trying to reduce fit-related returns, or an ML engineer who needs labeled body pose images to train a computer vision model, this is worth a free demo test. Also relevant if you're building animation or game pipelines that need markerless motion capture without a studio setup. Not the right fit yet if you need a vendor with proven enterprise customers, a public SLA, or documented ROI case studies — none of those exist publicly.
The free demo at demo.snapmeasureai.com costs nothing and tests in 10 minutes — the underlying probabilistic 3D reconstruction from two unconstrained photos is technically credible and aligns with what funded competitors like 3DLOOK have commercialized at $11M raised. The red flag is two years of near-zero public engagement: no named customers, no disclosed pricing, and three consecutive 1-upvote Product Hunt launches. Treat this as genuinely early-stage and do not build a production dependency on it without a contractual SLA.
Deep-dive insight, Easy and Pro modes, plus action playbooks — the full breakdown is one tap away.