Upload photos. Names handled automatically, included in all paid plans. Start your 14-day free trial.Get Started →

About

Whatwasbroken.Whatwebuilt.

Six problems sports photographers and athletic departments live with. Six things we did about each one.

GalleryID was built around the work photographers and athletic departments were already doing — and the unpaid hours they spent doing it. Each section below is a problem we lived with. What follows it is what we built in response.

Problem 01

Hundreds of photos. Zero names.

The game ends. The tagging starts. Saturday in October. A regional tournament weekend. The deliverable is named — but it took the next morning to get there.

Nobody pays for that hour. The deliverable does. The labor underneath it doesn't.

What we built
MatchID. The names attach themselves.

MatchID reads each frame, finds the faces and the numbers, matches them against your roster, and writes the names directly into the file.

The post-game hour disappears. The named photos are still there.

Problem 02

Names belong in the file.

Most platforms tag inside the platform. Export the file to Photo Mechanic, Lightroom, or a wire service — the names don't come with it.

The platform becomes the prison.

What we built
XMP and IPTC. Names travel with the photo.

GalleryID embeds athlete names directly in XMP and IPTC PersonInImage metadata on download. Open the file in Photo Mechanic and the names are already there. Send it to AP and the names go with it.

The file is yours. The names are in it.

Problem 03

Face recognition fails behind a mask.

Football helmets. Catcher's masks. Goalie visors. Ski goggles. Standard face recognition gives up on the toughest frame of the night — the action shot under the helmet.

That's where most identification tools stop.

What we built
MatchID reads more than the face.

Jersey numbers via OCR. Team context from the roster. Event context from the gallery. Most helmet-sport frames get named anyway.

The uncertain ones land in the review queue. No silent wrong answers.

Problem 04

A meter on identification taxes the season.

Most identification platforms charge 1 to 3 cents per image. Hundreds of photos per game becomes thousands by the playoffs, tens of thousands across a season.

The bill compounds as the season gets bigger. The harder you work, the more they take.

What we built
Unlimited identification. No meter.

MatchID is included on all paid plans. Storage scales with the archive. Identification doesn't have a meter.

$10 a month or $800 a month, same workflow, same review queue, same XMP.

Problem 05

Charging per seat penalizes the team.

$25 to $50 per user, per month. Add the photographer, the SID, the marketing lead, the design intern, the freelancer covering Saturday.

By the time the department is connected, the bill is built on people, not work.

What we built
Unlimited users. Roles control access.

Add the whole department. Add the freelance team. Add the booster who runs the social channel. Adding the next person doesn't change the bill.

Storage scales. Seats don't.

Problem 06

Your photos shouldn’t train someone else’s model.

A lot of AI tools improve by training on customer uploads. Your shots become part of their next release. Your athletes become part of their training set.

Their model gets sharper. You get nothing.

What we built
Trained on our own archive. Yours stays yours.

The base MatchID model was trained on our own photographic archive — hundreds of thousands of sports photos — before any client uploaded a frame.

Each client account refines from there, scoped to that account. The refinement doesn't flow back to our model and doesn't show up in another client's results. Delete the account, the learning goes with it.

Six problems. One platform.

Start Free

Match the name
to the face.

MatchID included on all paid plans.
Storage is the only line item.

Start 14-Day TrialView Pricing

No credit card required · 14-day free trial