Case Study: Adobe achieves fast, high-throughput complex queries for Typekit with Elastic (Elasticsearch)

Adobe - Customer Case Study

Typekit needed to serve complex, highly contextual font queries for its API, Browse UI and admin tools. The underlying data lived in many normalized MySQL tables and queries had to merge family metadata, per-font data and user-specific licensing context — all at high throughput and low latency.

The team denormalized records into Elasticsearch documents (MySQL remained source of truth) and syncs via an asynchronous after-save pipeline. Using term filters, queries and custom scripted queries to apply user context, they achieved near‑real‑time updates (seconds), simplified migrations and high performance — supporting thousands of QPS (5600 QPS reported on a 3-node cluster with 15 consumer nodes) for fast API and UI responses.

Adobe

Marko Iskander

Senior Computer Scientist