Everything on this page is functioning in the current build today. This isn't a vision deck: it's a walkthrough of the actual product, feature by feature, with real screen recordings and screenshots.
The core entry point of StartToFind. Customers search for products, shops, services, rooms, or categories and get real, mapped results back, with live counts, sponsored placements from local shops, and friend activity, not a generic web listing.
Beyond product search, customers can browse the directory itself, filtered by what actually matters to them, not just price.
The map layer that ties everything together, from city-wide search down to campus- and building-level detail. Wheelchair-, stroller-, and delivery-friendly routing is part of the same engine.
Deterministic, lightweight algorithms convert an uploaded floor plan into a pathfinding-ready graph: rooms, aisles, doors, and connections, editable and correctable in the browser. No expensive sensors, no heavy AI training required for the first version.
Build a list before visiting, then get it turned into a live in-store route with a synced checklist, nearby personalized offers, and budget tracking, down to a working scan-and-pay checkout.
Every participating business gets a step-by-step setup dashboard: building & layout, products & deals, scan & organise, shop info. A lightweight alternative to expensive e-commerce infrastructure.
Explainable predictive model
Buy advertising space, by intent
This is the engine behind the pilot's 200% ad-credit bonus and 3 years of free analytics: an explainable model that scores buying likelihood from real signals, without ever naming individual shoppers, only aggregated counts a shop can size and target advertising against.
It also reveals where shoppers dwell, where they stop and wait, which aisles they typically skip, and which locations are best for visibility actions. We do cross-correlation across the full product catalogue to see which products are bought together, so a shop can turn those patterns into better placements, bundles, and action plans.
Every mapped site doubles as an accessibility audit, flagging rooms unreachable by less mobile people, and scoring the building's overall connectivity and graph integrity.
Same app, same data, responsive down to a phone screen, because most local "before I leave home" searches happen on mobile.
Coming next
Being upfront about what's a working prototype vs. production-ready matters more to us than overselling a roadmap.
StartToFind Checkout works end-to-end today, but runs on prototype payments. Connecting a real payment service provider is next, before any shop takes live transactions through it.
The predictive model works from day one, but gets sharper with more signal. As pilot shops go live, targeting and forecasts improve with real shopper activity.
The responsive web app works well on a phone today; dedicated iOS/Android apps and embeddable widgets are the next step for reach.