Avleen & Sharan

The flagship photo site. 600 guests, four photographers, 14,000 photos across three days. The brief: every guest should find every photo they’re in within seconds. No scrolling through 14,000 thumbnails. No “which gallery again?”.

The pipeline:

• Photos upload to a private folder on the couple’s Google Drive
• A Python pipeline detects faces, clusters by identity, and writes face-cluster metadata to a Google Sheet
• The frontend reads the sheet at runtime and lets guests tap a face to find their own photos
• Video files get auto-thumbnails

The site is gated by passphrase, only invited guests get in. For this case study, we’ve embedded a video walk-through rather than a live link.