Semantic search with pgvector — distances, chunking and top-k queries

Python for AI Career is a Pro course

Choose between `<->`, `<=>` and `<#>`, and see why normalisation matters. Split documents into overlapping chunks, run top-k nearest-neighbour queries from Python, and use EXPLAIN to catch an index that is not being used.

The Python course is free and complete — this course is not made of anything taken out of it. Every chapter of a Pro course runs the full six steps:

• Encounter — Meet the problem before the definition — what cannot be done without this.
• Understand — Build the mental model: what Python actually does, and why this way.
• Worked — A complete example with its real output — run, not imagined.
• Predict — Say what happens before running it. A wrong prediction teaches more than a right one.
• Apply — Write it and run it on your own machine. Reading it is not the same as doing it.
• Stretch — Ten questions from easy to hard. The last ones are difficult on purpose.