Vespa AI Search Engine and Vector Database with Python

Discover Vespa architecture and how application packages, admin/config cluster, and stateless Java container cluster enable real-time processing, scalable search, and distributed query execution.

A tenant serves as a unique identifier and isolated space in Vespa cloud for your account or organization, organizing and securing configurations, data, and applications.

Discover the whisper cloud console overview, managing tenants and apps, accounts, billing plans, notifications, tokens, keys, secret store, and support within a trial environment.

Load and inspect the NF corpus from the hugging face dataset library; compare streaming and non-streaming, and note arrow data set and data set iterable types.

Create a Vespa cloud instance using interactive authentication by following the authentication link, confirming the device, and verifying a successful login, with certificates stored in Google Cloud files.

Create a vespa cloud instance using a pem file for automatic authentication. Upload and read the pem content, compare interactive authentication for development with key-based authentication for production.

Deploy a Vespa application to the Westpac cloud with a single line command, view deployment status, and retrieve the endpoint URL.

Feed documents to a Vespa application using Vespa format and feed iterable to index documents, with a callback monitoring success or errors and configuring the schema and namespace (tutorial).

Use the display hits function to turn spark query results into a pandas data frame for easy analysis and visualization of Vespa AI search results.

Explore plain keyword search in Vespa by indexing titles and bodies with BM25, ranking with tf-idf, and retrieving the top five results with YQL.

Explore Vespa hybrid search that fuses keyword search and semantic search with the or operator to retrieve documents matching either method, using reciprocal rank fusion and vector embeddings.

Explore hybrid search with the rank query operator that blends keyword search (BM25) and semantic search via vector embeddings with a reciprocal rank fusion ranking function.

Explore hybrid search with filters that combine keyword and semantic relevance to narrow results by date or document type. See queries converted to embeddings and ranked in a vector space.

Connect to a spark instance via the request library, retrieve a Vespa document, update its title, and verify the changes to maintain data integrity.

Reconnect the Vespa application by providing the certificate path, key path, and endpoint, uploading PEM files, and deploying to run a plain keyword search for verification.