SnapSeek

Inspiration

After waiting weeks for event photos, sifting through a 30 GB OneDrive folder from college and endless Google Drive links from a family wedding took hours just to find a handful of pictures. Scrolling through thousands of unindexed event photos is a universal headache for guests and a massive missed engagement opportunity for organizers.

I built SnapSeek.app to solve the needle-in-a-haystack problem of event photography using fast, automated AI face recognition.

What it does

SnapSeek is an AI-powered event photo-sharing platform designed for photographers, event organizers, and wedding hosts to distribute photos instantly.

• Frictionless Guest Discovery: Guests scan an event QR code, snap a quick selfie, and immediately view a private, curated album containing every photo they appear in.
• Massive Gallery Indexing: Instantly searches across thousands of high-resolution event photos in milliseconds.
• Zero-Friction Access: 100% browser-based. Guests do not need to download an app or create an account to get their pictures.
• Host & Photographer Controls: Fast batch uploads with simple album management and private sharing controls.

How I built it

• Frontend: Responsive, lightweight modern web interface optimized for seamless mobile-first guest access with zero app install requirements.
• Client-Side Optimization: Implemented client-side image compression to drastically reduce upload payloads, speed up ingestion, and preserve bandwidth.
• AI & Face Recognition Pipeline: Integrated deep-learning face detection and vector embedding models to index and match attendee faces with 99.9% accuracy, even across varying lighting, angles, and crowd backgrounds.
• Serverless Backend: Built on a scalable cloud architecture with asynchronous processing queues and vector search to handle high-concurrency event uploads and real-time query matching.

Challenges I ran into

• Handling Massive Batch Uploads: Ingesting thousands of high-resolution RAW/JPEG files simultaneously without throttling server resources required tuning asynchronous background workers and client-side pre-processing.
• Precision in Dynamic Environments: Tuning facial recognition thresholds to reliably identify attendees across candid angles, low-light wedding receptions, and sunglasses without returning false positives from background crowds.

Accomplishments that I'm proud of

• 100% Web-Native: Eliminated app store friction entirely—guests go from QR scan to selfie match in under 5 seconds.
• Sub-Second Search Latency: Real-time vector search across thousands of gallery assets.
• High Precision: Reached 99.9% AI facial matching accuracy across diverse event lighting and group compositions.
• True End-to-End Automation: Turned a multi-hour manual sorting chore into an instantaneous automated workflow.

What's next for SnapSeek

• Expanding real-time live photo streaming during active events.
• Adding customizable branding, watermarking, and download analytics for professional event photographers.
• Introducing automated high-res batch delivery integrations and multi-event organizer portals.

Built With

• eventphotos
• photosharing
• web