AI face recognition

Face Recognition That Finds
Every Guest in the Crowd

One selfie is all a guest gives you. Face search finds every photo they're in, group shots included, while you're still shooting, and the ones from before they scanned.

[ 01 of 07 ]
How a match happens

From Shutter to the Right Guest in Four Steps

No tagging, no sorting, no folders. The only thing a guest gives you is one selfie at the start of the night.

  1. 01

    Find every face

    The capture app spots each face as the photo comes in: in crowds, at an angle, in a dim hall.

  2. 02

    Make a face print

    Each face becomes a face print, a list of numbers that describes it. Prints are compared, not photos.

    • 98%
    • 41%
    • 23%
    03

    Compare with every selfie

    Every print is checked against each guest's selfie for this event. Only strong matches count.

  3. Priya+1
    New
    04

    Deliver it live

    The photo lands in the gallery of every guest it matched, in seconds, while you keep shooting.

[ 02 of 07 ]
Every guest, every photo

Group Shots and Late Arrivals, Handled

The two places manual sorting always fails: the twelve-person group photo, and the cousin who scans at the end of the night.

Group photos

Every face in every frame

The capture app finds each face as the photo comes in and turns it into a face print. A group photo with twelve guests lands in twelve galleries.

  • Faces read on the phone as you shoot
  • No limit on faces per photo
  • Built for crowds, angles and low light
Late scans

Scan late, still get everything

A guest who scans the board at 21:52 is matched against everything you shot since the mehndi. Photos you upload later from home are matched against every selfie too.

  • Backfill starts the moment a selfie arrives
  • Late uploads re-match on their own
  • One selfie per guest, for the whole event
Match time
Under 10 s
Faces per photo
No limit
Selfies per guest
1
Re-matching
Automatic
[ 03 of 07 ]
Built for Indian weddings

Real Rooms, Not Studio Headshots

Where it shines, and the two places a camera simply can't see a face. We'd rather tell you than have a guest find out.

  • Matches

    Dark halls and sangeet lights

    Coloured lights and dim halls are the normal case, not the exception. The matching was built for them.

  • Matches

    Crowds and group photos

    Every face in a twenty-person group photo is matched on its own, so each guest gets the shot.

  • Matches

    Side angles and candids

    Guests rarely look at the camera. Profiles and three-quarter turns still find their match.

  • Matches

    Haldi paste and gulal

    Turmeric on the cheeks or a smudge of gulal doesn't hide who someone is.

  • Know the limit

    Sehra, ghunghat and veils

    When a sehra or ghunghat covers the face, there is nothing to match. Those photos stay in the full event gallery.

  • Know the limit

    Turbans, caps and sunglasses

    Turbans and caps are fine. Big sunglasses can hide too much, and the next photo usually finds them.

[ 04 of 07 ]
Privacy

Privacy, Plainly

Guests are trusting you with their face. Here is exactly what happens to it, in words they can read before they take the selfie.

  • One selfie, one purpose

    A guest's selfie is used to find them in that event's photos. Nothing else.

  • Only their own photos

    Each guest's gallery holds the photos they are in. Nobody browses anyone else's.

  • Deleted when they ask

    Guests can ask for their selfie and match data to be deleted at any time.

[ 05 of 07 ]

Hand Sorting vs Face Recognition

How event photos usually reach guests, next to how they reach them with ShotKite.

Sorting by hand
  • A week of sorting folders by face
  • Group photos sent to one person, if at all
  • Guests scrolling 3,000 photos to find themselves
  • Late arrivals miss everything before they came
  • One shared link anyone can forward
ShotKite
  • Every photo matched while you shoot
  • Each group photo reaches every guest in it
  • A gallery of only their photos
  • Late scans backfill the whole event
  • Each guest sees only their own photos
[ 06 of 07 ]
Works with

What the Matching Powers

Live galleries for guests, album selection for the couple, and photos straight from your camera.

[ 07 of 07 ]
Questions

Face Recognition Questions, Answered

How does face recognition photo sharing work?

Each guest scans your QR and takes one selfie. As you shoot, every face in every photo is compared with those selfies, and each photo is added to the gallery of every guest it matches, live.

How accurate is it at a wedding?

It was built for dark, crowded, fast-moving rooms rather than studio headshots, and only strong matches are delivered. Instead of a percentage on a website, you get the match rate for every one of your events on the dashboard.

Does it work in the dark or under coloured lights?

Yes. Sangeet lights and dim halls are the normal case it was built for. A very dark or blurred selfie is harder to match, so good light at the QR board helps.

Will guests get the group photos they're in?

Yes. Every face in a photo is matched on its own, so a group photo of twelve guests lands in all twelve galleries.

If a guest scans late, do they get the earlier photos?

Yes. A new selfie is checked against everything already shot at the event, so someone who scans at 10 pm still gets the photos from the start of the evening.

Is the selfie stored? Can other guests see my photos?

The selfie is used only to find the guest in that event's photos, and they can ask for it to be deleted. Each gallery is private to its guest: nobody sees anyone else's photos.

Free plan · No credit card · 5-minute setup

One Selfie. Every Photo
They're In.

Turn it on for your next event. Guests stop asking you for photos, because they already have them.

Early access

Join the ShotKite waitlist

We’re opening access in batches. Get an invite first.