Searching movies semantically on Chromia

You will build a full-stack app that turns movie plot summaries into vector embeddings and stores them on Chromia. The full movie metadata lives on-chain, and the embeddings are indexed with the vector_db_extension. With this in place, you can run semantic searches that match queries by meaning and return detailed results straight from the blockchain.

Key learning objectives

• Prepare your project environment and run the backend on the Chromia testnet
• Create vector embeddings and upload them together with movie metadata
• Find semantically similar movies with natural-language queries and get detailed results back

The course is hands-on: you work with the Carnegie Mellon University (CMU) Movie Summary Corpus and Python-based tools to learn how semantic search works on-chain.

Repository link

The full source code for this course is available in the Vector DB Movie Demo repository.