Vector Databases Workshop

This workshop is for Python developers who want to store vector embeddings in databases. It teaches you how to

• recommend a vector store to use in different situations
• use a local and a cloud vector store

Pre-requisites: Familiar with embeddings and Python

How to pick a vector store

1. Use the client's cloud environment's vector database. Pick PostgreSQL (pgvector) or MongoDB based on developer preference.
2. Else use ChromaDB locally.

If performance is a factor, use a specialized vector database.

See Databases in the Skills Radar for a full list.

Exercise

• Download the workshop Jupyter notebook.
• Upload your notebook to Colab.
• Run the notebook as-is.
• Add a new vector database like sqlite-vss, faiss, scikit-learn, etc. Run the same exercise with this vector database.

Submit your code

1. Create an issue titled Exercise submission. Add a link to your notebook

Verification

To mark a submission as correct:

1. Check if a new vector database was added.
2. Check if the code has been fully executed with these changes without errors.

Video