Case Study: Oyster achieves 95% faster image processing (1M photos in 20 hours) with Amazon Web Services

Oyster - Customer Case Study

Oyster.com is a New York–based hotel review site known for detailed, investigator‑shot photo coverage—hundreds of images per property and more than one million images overall. Storing and routinely reprocessing that volume in‑house became costly and slow: resizing and watermarking runs took hundreds of hours, new device formats (like Retina iPad) added long delays, and buying extra hardware was expensive and inefficient.

Oyster moved its images to Amazon S3 and shifted processing to Amazon EC2 (using a custom Linux AMI, ImageMagick, boto, SQS and EC2 Spot Instances) to scale batch jobs up to ~100 instances. The change cut full‑collection processing from hundreds of hours to about 20 hours (≈95% faster), reduced time to produce Retina images from about a week to a few hours, and saved roughly $10,000 in capital and $10,000 in operating expenses while freeing teams from hardware maintenance.

Oyster

Eytan Seidman

Co-Founder and Vice President of Product