Patent Embedding Radar 2D: Spotting Competitor Encroachment
Interactive 2D embedding-space radar for AI-driven competitive intelligence: patent filings plotted as directions on a similarity circle, with live cosine-similarity encroachment detection and whitespace-gap discovery — drag to pan, scroll to zoom.
AI-driven competitive intelligence turns patent filings, product pages and marketing text into vectors in an embedding space, then measures similarity between them to spot threats and gaps automatically. This 2D simulator makes that pipeline visible: filings from your company and three competitors stream in continuously and settle onto a similarity circle according to which of six technology clusters they belong to. A live cosine-similarity check flags any competitor filing that lands too close to your own portfolio — exactly the nearest-neighbour math a real patent-monitoring system runs — while a separate probe search continuously hunts for the circle's most isolated arc, the whitespace a real R&D team could stake out uncontested. Drag to pan the view and scroll to zoom into a crowded cluster.
A flat 2D embedding-space radar for AI-driven competitive intelligence: patent filings from your company and three competitors stream in and settle onto a similarity circle, with live cosine-similarity encroachment detection and a probe search for the largest unclaimed whitespace arc, drag-to-pan and scroll-to-zoom included.
2D · HTML5 Canvas 2D · 60 FPS target · runs fully client-side, no install