OnPoint is a curator-first styling and retail-intelligence platform. Curators — small-business fashion sellers, often WhatsApp-based in emerging markets — run branded storefronts at /s/[slug]; their customers get AI styling, virtual try-on,and checkout via M-Pesa and a self-custodial agent wallet. Every session generates live retail intelligence: product gaps, competitor pricing, demand signals, and recommended GTM actions.
Core capabilities:
OnPoint's mission is to give every fashion curator a branded, AI-powered storefront and every shopper a confident styling decision — bridging inspiration, try-on, and purchase across web, mobile, and WhatsApp, with verifiable on-chain commerce.
Fashion retail in emerging markets runs on WhatsApp. Small-business sellers (curators) hand-pick stock, photograph it, and resell it through chat — but the workflow is entirely manual: a customer messages, the curator describes items by hand, arranges payment, and tracks orders in their head. There is no branded storefront, no try-on, and no automated intelligence on what to stock next or how to price it. End-shoppers, meanwhile, get outfit inspiration scattered across Pinterest, Instagram, and TikTok with no way to visualize, personalize, or confidentlyact on it — and no path from "I love this look" to actually buying or owning it
OnPoint turns every curator into a branded storefront with an AI styling layer. A curator onboards in 30 seconds at /curator/onboard and receives /s/[slug] — branded listings, AI try-on, polaroid share cards, M-Pesa STK push checkout, and WhatsApp receipt sending. Their customers point a camera at an outfit, get instant AI styling feedback and matching products, and buy in-flow. When the curator'scatalog lacks a match, OnPoint's agent browses the open web, compares market options, and surfaces product gaps, competitor pricing, and recommended actions. A persistent autonomous worker (PM2) polls trending fashion hourly, tracks price history per item, detects drops, and auto-buys within user-set spending limits — recording verifiable on-chain receipts for every autonomous action.
