ParallelDots builds computer vision for retail execution. Its flagship ShelfWatch analyses photographs of store shelves to detect SKUs, measure share of shelf, check planogram adherence, verify price tags and promotions, track competitor products and flag out-of-stocks. The site claims up to 98% accuracy on captured images, results within 60 seconds and on-device recognition that works offline, and it adds Scan and Go, an augmented-reality audit flow, and a fixed shelf-mounted ShelfWatch Camera.
Newer products address the hard part of shelf recognition, which is new SKUs: Saarthi is described as training models to detect new SKUs within 48 hours, and SmartGaze is an eye-tracking research product. The customers listed are CPG manufacturers such as Mondelez, Unilever, Heineken, BAT, Reckitt and ITC, and the site reports operation in more than 50 countries, 2.5 million-plus outlets and 15 million-plus photos processed monthly. Case results such as Unilever Ghana’s on-shelf availability rising from 46% to nearly 91% are vendor-reported.
Pricing is quote-only and no trial is mentioned. Integrations with sales-force automation, CRM and ERP systems and with field audit agencies are the practical questions, and recognition accuracy in your own categories and store formats, and how fast new packaging is learned, decide the value more than the headline accuracy figure.
