Vespa

What is Vespa?

Vespa is an open-source data platform that combines big data handling with artificial intelligence, letting businesses apply AI directly to their data in real time. It is built to scale and offers high performance for companies that want to get more value from their data.

With Vespa, you can build advanced search engines, recommendation systems, and personalized content delivery, all while processing information in real time. Its data processing and machine learning capabilities support faster, better-informed decisions. The platform is designed for high availability, so services stay up and running, and it lets you create applications that blend fast vector searches and filtering with up-to-date machine learning models. You can use the cloud service or host Vespa yourself.

Who created Vespa?

Vespa was created by Jon Bratseth, Kim Johansen, Frode Lundgren, and Kristian Aune. Jon Bratseth serves as CEO and founder and has more than 20 years of experience designing and programming large, distributed systems. The company is based in Trondheim, Norway, and develops the open-source Vespa platform, combining big data handling with AI for advanced search engines, recommendation systems, and real-time AI applications at scale.

What is Vespa used for?

• Generative AI: Vespa helps power generative AI applications, merging vector, text, and structured data searches and using machine learning for relevance and tensor computations.
• Search: It is a complete search engine and vector database. You can perform vector searches, keyword searches, and searches over structured data in a single query, with built-in machine learning for real-time AI.
• Recommendations and personalization: Build recommendation and personalization systems that combine fast vector searches and filtering with machine-learned models that evaluate items to tailor experiences.
• Semi-structured navigation: Group data dynamically for browsing and filtering, working alongside search and recommendation features.

Key features

• Scalable big data: Process and apply AI to large volumes of data.
• Machine learning integration: Integrate machine learning models for real-time data inference.
• Versatile search: Vector, keyword, and structured data searches within a single query.
• Real-time performance: High-speed search and personalization delivered in real time.
• High availability: A proven architecture that keeps applications running consistently.

Getting started

You have two main options: download the open-source software or use the Vespa cloud service. Once set up, you can build applications that combine fast vector search with machine-learned models for search, recommendation, and personalization use cases.