Philips had a not-so-small problem: a bloated image library of 200,000 photos and a global marketing team struggling to find anything in it. Manually sorting images by file names and outdated metadata wasn’t cutting it anymore. So they brought in Google Cloud’s Vertex AI and a generous helping of computer vision to solve it. The result? The mess was narrowed down to a much leaner, more brand-consistent 8,000 images—processed in hours, not weeks.
This wasn’t just tagging photos with prettier labels. Philips’ team built an algorithm using image embeddings—basically, machine-parsed visual fingerprints that ignore wonky file names and focus on content. The AI grouped visually similar photos, selected the strongest “master” versions, and tossed the duplicates. Since a cropped or compressed image still shares the same essential visual DNA, metadata didn’t even come into play. According to Philips’ marketing lead, this metadata-free system delivered faster, more accurate results while freeing up human teams to focus on storytelling instead of file sorting.
Beyond cleaning up the backlog, the new system helps with compliance, version control, and campaign agility. For a company with a diverse product set and strict brand guardrails, that’s more than just tidy cabinets—it’s operational breathing room. The AI also paves the way for faster A/B testing and more responsive campaign rollouts. Philips isn’t just keeping up with content velocity—it’s finally in the driver’s seat.
Read more at PYMNTS.