Cassandra

A field of outlined movie record cards, four larger cards linked by orange lines with a matching orange word segment, next to the text "cMovie - Week 2 - One FastAPI endpoint, five filters"

Not every search needs Elasticsearch: build a Python FastAPI filter and keyword search on Cassandra

SAI matches whole values, not fragments. Split titles and cast names into words at load time, index the words, and serve fragment search for 1,000 real movies from a small Python FastAPI app.

A dense field of flat tiles thinning toward the left, with three ranked results picked out in orange, next to the text "A 10-week series - Search by what happens, not the title"

Build an AI movie recommender Python and TypeScript app on Cassandra

Ten weekly posts that build one AI movie app, cMovie, from scratch on Apache Cassandra 5.0 - SAI filtering, vector search and a RAG assistant, one working part at a time.

A field of grey, crossed-out column tiles thinning into connected orange tiles, next to the text "cMovie - Week 1 - Stop typing ALLOW FILTERING"

Query any column, not just the primary key: a SQL developer's guide to Cassandra SAI

Cassandra only answers queries on the primary key until you index the other columns. See how Storage Attached Indexes lift that restriction, using 1,000 real films.

You already trust this DB, let it run your AI stack too

Bolting a separate vector database onto your operational store creates two schemas, two consistency models, and two failure domains. See why Cassandra keeps it all in one system.

You picked the wrong database

Not all AI databases are created equal. See how Cassandra runs vector + metadata search in a single query, at scale.