When I was in China, I worked as a video director and photo director for a design agency, where I handled their video and photo. I quickly realized their photos were a mess. Different naming conventions in different folders. Different formats, duplicates, no metadata, no standard. And no DAM system to find their photos. That’s where the idea of using AI to clean things up came from.
A universal problem
What I saw over there, I keep seeing everywhere since. In architecture firms, communication agencies, design studios, heritage institutions. No matter the sector, the pattern is always the same: years of visual production piled up without discipline. A project, a folder, a file name that only makes sense to the person who created it — and even then, only on that day. Six months later, no one knows where the right version is, which one is rights-cleared, which one was taken by whom, when, where.
It’s as if photography, unlike contracts or finances, never earned its own administrative rigor. It’s treated as a byproduct of the project, not as an asset in its own right. The result: teams paying again to redo photos they already have, designers losing hours digging through hard drives, art directors using an image without knowing if it’s still under license.
Why metadata isn’t a technical detail
Well-made metadata is the memory of a photo. Who took it, when, where, under what usage rights, for which client, in what context. Without it, every image becomes a puzzle that has to be pieced back together by hand — when that’s even still possible.
This is even truer today because the context has changed. With the rise of AI-generated or AI-edited images, the question is no longer just “where is my photo,” it’s “is this photo authentic, and can we prove it.” A company, an agency, an architecture firm that publishes images without traceability now faces a real credibility risk — from a client, the press, or simply a platform algorithm that’s starting to filter unverified content.
C2PA: provenance as the new standard
This is where C2PA (Coalition for Content Provenance and Authenticity) comes in — a standard that attaches a verifiable history to an image: where it came from, what tools it passed through, whether it was edited or AI-generated. This is no longer a luxury reserved for press agencies. As trust in images becomes a market issue — for photojournalists, but also for companies publishing visuals of projects, products, buildings — C2PA becomes as much a selling point as a legal safeguard.
An architecture firm that can prove its renderings and construction site photos are authentic stands out. A company that can find any image in its library in seconds, with the right usage rights attached, saves time and avoids costly mistakes. Digital cleanup is no longer a documentation chore: it’s a competitive asset.
Cleaning up, at human scale
The real problem is that no one has time to manually reorganize tens of thousands of photos. That’s exactly the wall I ran into: everyone knows it needs to be done, but it feels like a never-ending project.
That’s why I started developing tools that use AI to automate this cleanup — generating IPTC/XMP metadata, structuring file naming, applying C2PA provenance, without requiring weeks of manual work. The idea isn’t to replace human judgment, but to give it back the time it needs for what actually matters: choosing the right image, not spending an hour looking for it.
DIGITUM Consulting
That’s how DIGITUM Consulting came to be. We’ve developed DIGITUM Entreprise to address this exact problem — whether you’re an agency, an architecture firm, a design studio, or any other organization that produces or manages large volumes of images. The idea is simple: bring order to visual chaos, without spending your life doing it.
Contact us for a demo.
If your image library looks like what I found several years ago, it doesn’t have to stay that way. It’s just a cleanup that hasn’t been automated yet.