Case Study: Sajari achieves sub-10ms search latency and global scale with Google Cloud Platform

Sajari Powering expansion of search and match business with Google Cloud

Sajari, an Australian search-and-matching technology company founded in 2013, provides APIs that let businesses add machine-learning powered search and matching to websites and systems. As the business grew to hundreds of customers and roughly 1 billion requests per month, inconsistent performance and network latency from its original cloud provider threatened its ability to return complex matching queries (often analyzing millions of data points) within the sub‑second response times customers expected.

Sajari staged a migration to Google Cloud Platform—using App Engine, Cloud Datastore, BigQuery, Compute Engine and ultimately Google Kubernetes Engine—taking advantage of low-latency networking, consistent compute, live migration and advanced load balancing. The move delivered a median search query time under 10 ms, match times within business SLAs, high availability, faster scaling across geographies, and freed engineers to focus on product rather than infrastructure.

Sajari

David Howden

Senior Engineer