Vespa
Vespa is an open-source big data serving engine designed for low-latency queries over massive, constantly changing datasets. It uniquely combines capabilities for full-text search, structured data queries, vector search, and real-time machine-learned model evaluation into a single, scalable, and…
Key facts
- First appeared
- 2007
- Category
- database
- Problem solved
- Vespa was created to solve the problem of serving extremely low-latency, personalized, and machine-learned ranked queries over massive, constantly changing datasets, integrating full-text search, structured data, and real-time ML inference into a single, scalable, and fault-tolerant engine. Its predecessors or alternative approaches often involved complex, brittle, and higher-latency pipelines combining separate search engines, databases, and offline ML prediction systems.
- Platforms
- Cloud (AWS, GCP, Azure), Linux, Kubernetes, Docker
Related technologies
Notable users
- Yahoo! / Verizon Media
- eBay
- Workday
- Squarespace
- Spotify