From memory card to searchable gallery
Bibbit sits between your gallery host and your participants: it reads every bib, groups every face, lets you approve the results, and writes them back as keywords runners can search. Here's each step up close.
Connections
Connect the galleries you already use
Bibbit works where your photos already live. Link SmugMug with one click of OAuth, or sign in with your Zenfolio account — and connect as many accounts of each as you shoot with. Your photos never move; Bibbit only reads them and writes keywords.
- ✓SmugMug and Zenfolio, multiple accounts per host
- ✓Credentials stored encrypted; access revocable at any time
SmugMug
Connected as Northside Race Photos
Zenfolio
Connected as northsideraces
Roster
A simple CSV turns numbers into names
Upload the event's roster — two columns, bib and name — and every detected number is matched against it. Detections stop being digits and become runners: keywords carry both the bib and the name, so participants can search by either. And when a face cluster is resolved, the roster names those photos too — including the ones where no bib is visible.
- ✓Bib number and runner name both become searchable keywords
- ✓No roster? Bib numbers alone still work
roster.csv
bib,name
1247,Maya Ruiz
882,Jon Okafor
451,Aiko Tanaka
3 runners matched
Bib detection
Every photo scanned, every bib read
Pick the albums for your event and hit Scan. Bibbit works through every photo in the background — detecting bib plates with a model trained on race photos, reading the numbers, and matching them against your roster CSV so “1247” becomes a runner with a name.
- ✓Roster matching turns numbers into names
- ✓Photos are fetched, processed, and discarded — never stored
Spring City Marathon — Finish Line
Scanning1,147 / 1,852 photos
Spring City Marathon — Course km 32
CompleteFace grouping
Find runners even when the bib is hidden
A pinned arm or a jacket zip can hide a bib for half the course. Bibbit groups photos of the same runner by facial similarity, so resolving one cluster to a bib number tags every photo that runner appears in — visible bib or not.
- ✓One answer tags a whole cluster of photos
- ✓Face data stays in your account and is deleted with the scan
1247 · Maya Ruiz · 18 photos
ResolvedReview
You approve everything before it goes live
AI results are probabilistic; your gallery is not. The Review tab walks you photo by photo — detected bibs, matched names, and boxes drawn on the image — so you approve, fix, or reject each result. Nothing reaches your gallery until you say so.
- ✓Fast photo-by-photo review, built for thousands of frames
- ✓Manual tagging and drawing for the tricky ones
Detected bibs
1247
Maya Ruiz
882
Jon Okafor
Sync & export
Approved keywords, written back home
One sync pushes approved bibs and runner names to your gallery as searchable keywords — on demand, or automatically as you review. Only changed photos are written, existing keywords are preserved, and everything is also downloadable as CSV or JSON.
- ✓Manual by default; opt-in auto-sync per gallery
- ✓CSV/JSON export mirrors exactly what sync writes
✓ Synced · 213 / 213
Syncfinish-line_0417.jpg, "1247, Maya Ruiz"
Download CSVThe payoff
Runners search a number, find themselves
The finish line of the whole pipeline: a participant opens your gallery, types their bib number into the search they already know, and every photo they appear in comes back — including the ones a face match found. More photos found means more photos sold.
- ✓Works in your host's own gallery search — no new app
- ✓Face-matched photos show up too, not just visible bibs
34 photos found
See it on your own gallery
Connect an account, scan an album, and watch the first bibs come back in minutes.