Welcome

In "From data to agent" you read the report that an AI agent handed to the management of Hornos Lumbre. Now you are going to redo its calculations. The managing director wants to see the queries behind each figure, and it falls to you to write them with your copilot, read them before running them and check what they return. Keep the case data at hand, the same seven files, and your AI assistant: in the first lesson you set up with them the workbench you will use throughout the pathway. You do not need to know how to program. You need not to accept any figure until you have counted the rows.

What you will find

• Block 1. The workbench and the table. Where what your copilot writes runs and how to read code you have not written. And the unit of everything else: the table, with its grain and its key. The risk you learn to see: accepting a table without knowing what a row is or how many it should have.
• Block 2. Asking: select, filter and aggregate. A query reads like a sentence: from which table, which rows, grouped how and counting what. And the usual stumbles when summing, counting and averaging. The risk you learn to see: the filter in the wrong place and the count or average that does not measure what its name says.
• Block 3. Joining tables without duplicating rows. Joining two tables is matching rows by a key. When a row finds two matches it is repeated and the sums inflate, with no error or warning. You learn to read a join, to count it before and after and to fix it in two ways, and with that you redo step 6 of the trace: the join that inflated the promotions.
• Block 4. The copilot that writes the query. What a copilot needs to get a query right, how to check the one it delivers and what permissions it is given over a database. You write the definitions that the agent decided on its own, commission sales by region and hand in the queries for points 1, 2 and 3 of the report, each with its record.

How you will work

Each lesson opens with a short introduction and, below it, several pills: short pieces that explain a single idea and end with a case or something to do. Read them in order the first time; afterwards you can go back to the one you need.

Where you see an interactive practice, do it before reading on: it is there for you to decide, calculate or classify, and the explanation comes afterwards. It does not save what you write, so copy whatever you want to keep.

Each block ends with an exam of six questions. All the answers come with a comment, including the wrong ones: failing and reading why is part of the method.

At The Singular Academy you learn by solving real cases. If at the end of a lesson you do not know what you would do differently tomorrow, go back over it.