Case Study: Artfinder achieves rapid AI-powered art recommendations and 75% revenue growth with Amazon Web Services

How Artfinder Powers Art-Matching Services Using AWS

Artfinder is an online marketplace that helps thousands of independent artists sell directly to buyers. Facing incomplete artist-supplied metadata and growing traffic, the company needed a scalable way to analyze and match artwork to buyer tastes and to collect detailed behavioral data without overloading its transactional systems.

Using AWS services such as Amazon Rekognition, Amazon Machine Learning, Amazon Kinesis Firehose, Amazon EC2, Amazon Redshift, and Amazon CloudFront, Artfinder built image-analysis and recommendation pipelines—prototyping in four hours and moving to production within a week—and ran large offline analyses in days instead of weeks. The result: a better matching experience (including a recommendation Twitter bot), rapid feature launches, improved global performance, and strong business impact—revenue up 75% in the U.K. (190% in the U.S.) and 60% of traffic from outside the U.K.

Artfinder

David Tilleyshort

Chief Technology Officer