What is SQL?

SQL (Structured Query Language) is how you ask a relational database for rows — filter, sort, and combine tables with a declarative query instead of hand-written loops.

Executive summary

Treat SQL as the language of tabular truth: customers, orders, events. AI systems still need clean joins and filters before embeddings or agents touch the data.

Watch

Video coming soon

Learn

A database stores data. A table is a named grid of rows and columns. A query asks for a result set.

-- Conceptual shape (dialect details vary)
SELECT column_list
FROM table_name
WHERE condition;

Why this matters for AI work:

• Training and eval sets often start as SQL extracts.
• RAG and analytics both need trustworthy filters (WHERE) before “smart” layers.
• You debug pipeline issues faster when you can read the source tables.

Check

Check understanding

What is a SQL table?

• A named grid of rows and columns
• A trained neural network
• A folder of PDF files

Which statement is true?

• SQL is only useful after you have an LLM
• SQL lets you declaratively read and filter tabular data
• SQL replaces the need for data modelling

What to do next

1. Continue to SELECT, FROM, WHERE.
2. Skim Data and knowledge when designing extracts for AI.