Weaviate uses GraphQL to provide user-friendly data interaction. Weaviate is an open-source vector search engine, and all searches (e.g. semantic, contextual) are done via its GraphQL API. We’ve put a lot of thought into the design of the GraphQL API, which results in good user and developer experience. In this talk, I will take you along in the journey of how our GraphQL implementation was shaped according to user needs and software requirements, and show a demo of the current design for Weaviate. The demo will show how Weaviate’s GraphQL design enables semantic (vector) search in combination with scalar search through unstructured data. Machine learning models are used in the background, but with the current GraphQL design, users without a technical background can query the vector database easily.
Weaviate has a modular architecture, so users can connect various machine learning models on top of the vector database. Examples are the newly released Question Answering module and the Named Entity Recognition module. Modules can extend the GraphQL schema dynamically, to query the new features intuitively.
This presentation contains a demo where we will query the complete Wikipedia, conduct semantic search queries and more. All through Weaviate’s GraphQL API. No prior knowledge is required.