Case Study

QuestDB enables machine learning engines that power Yahoo search

Yahoo use QuestDB in an embedded capacity within their machine learning engine. This system intelligently services close to a billion users at a rate of 500k queries per second.

Massive performance
Billions of users, massive queries per-second. No problem.
ML + hyper ingest
ML engines require the latest, fastest data in their pipelines. QuestDB delivers.
Fault Tolerance
Reliable data monitoring at massive scale, where every impression counts.
An advertisement for yahoo.com showing personalized search across multiple mobile devices
Avg ingested rows/sec
3M+
Write speed vs InfluxDB
10x
Compression ratio
6x
Cloud up-time
99.99999%
Chart showing a continuous integration pipeline for Yahoo's Vespa engine

Vespa Machine Learning Engine

Yahoo's Vespa Platform

Yahoo relies on a custom machine learning engine that powers search and recommendation, serving personalized content to hundreds of millions of users in real-time. For high-quality insights out, ML engines need high-quality data in. QuestDB deduplication and out-of-order indexing keep the streams blasting.

Efficient Metrics
Monitor metrics for autoscaling decisions in real-time

QuestDB + ML

Intelligent data monitoring at true scale

Yahoo stores and analyze application monitoring metrics in real-time.

Chart showing resource utilization of nodes within Yahoo's Vespa engine
Cost-effective monitoring
No external monitoring solutions required for autoscaling decisions
Embedded Analytics
Store and analyze application monitoring metrics quickly within the application
Native Time Series
QuestDB enables native timeseries support within the ML engine

"We use QuestDB to monitor metrics for autoscaling decisions within our ML engine that provides search, recommendation, and personalization via models and aggregations on continuously changing data."

Jon Bratseth
VP Architect at Yahoo

Advanced Machine Learning

QuestDB in your modern stack

Yahoo leverages QuestDB's high-performance and fault tolerance to power its custom machine learning engine for real-time analytics. Massive ingestion in, with familiar SQL for data-out. Dashboards, queries, it's yours in moments.

Chart showing resource utilization of nodes within Yahoo's Vespa engine

Ready to upgrade?

Break free from ingestion speed bottlenecks

Spin up in minutes

Download QuestDB