Bring AI-Based Search to Your Web App – Sebastian Witalec, JSNation 2023

Discover how to integrate AI-powered search into your web app using Weaviate, CLIP, Cohere, and more. Learn about vector embeddings, semantic search, and multilingual models.

Key takeaways
  • Machine learning (ML) has become accessible, allowing developers to bring AI-based search to their web applications without needing a Ph.D.
  • Weaviate is an open-source vector database that can be used to store and search data using ML models.
  • Vector embeddings are used to represent data in a multidimensional space, allowing for semantic search.
  • The CLIP model can be used to understand images and text, enabling image search and multimodal search.
  • The Cohere API can be used to generate text and perform other NLP tasks, such as summarization and translation.
  • The builder pattern can be used to create complex queries in a clear and concise way.
  • GraphQL can be used to query data from Weaviate in a flexible and efficient manner.
  • Multilingual models can be used to perform search in multiple languages.
  • Generative models can be used to generate new content, such as text, images, and code.
  • Vector search can be used to find similar objects in a database based on their vector representations.
  • Semantic search can be used to find relevant results based on the meaning of the query, rather than just keyword matches.