Structured Output: Tables, Lists and JSON on Request
Get answers in a fixed shape you can paste into a spreadsheet, a form or another tool.
A taste of a lesson
I asked for a table of tasks from my meeting notes and it gave every task an owner, but half of those were never mentioned.
That is a classic: when every cell looks like it needs filling, the model fills it with plausible guesses. Add two rules to your prompt. First: 'If the owner or date is not stated in the notes, write unknown.' Second: add a column called source_quote with the exact words from the notes for each task. Now any owner without a matching quote stands out immediately. Try rerunning it on the same notes. How many rows now show unknown for the owner?
Written by the teacher as an example. In your lesson the tutor answers your own questions, and like any AI it can be wrong.
What you will be able to do
- Define fields, allowed values and missing value rules before prompting
- Request clean tables, CSV and simple JSON from an assistant
- Check structured output for invented values and dropped items
- Choose the right format for a spreadsheet, a form or another program
Lesson plan
- 1 Start from the destination Pick a format based on where the output goes next. Start
- 2 Fields, templates and allowed values Define the shape so precisely that the model has little room to improvise. Start
- 3 Missing information and invented values Stop the model from filling gaps with plausible guesses. Start
- 4 JSON without fear Read and request simple JSON confidently, even without coding experience. Start
- 5 Long lists and dates Handle long extractions and date formats without silent errors. Start
- 6 A reusable extraction prompt Build one dependable prompt for a recurring extraction task. Start
Try asking
About this tutor
Free text is fine for reading, but much work needs data in a fixed shape: a table of action items, a list with the same fields every time, or JSON another program can read. This tutor teaches you to request structured output reliably: defining fields, giving a template, handling missing values, and checking the result before you trust it. You practise extracting details from emails, notes and documents into tables and simple JSON, and learn why chat tools sometimes produce broken formats or invent values to fill a field. No programming is required, though it helps if you have seen a spreadsheet or a JSON file before.
Reviews
Students can review a tutor after a paid lesson. Nobody has yet.
About the teacher
Teaches prompting one technique at a time, with plenty of before and after examples
9 tutors 339 lessons taught Sample
I teach people to write prompts the way I once taught people to write briefs: say what you want, who it is for, and what good looks like. I came to AI from editing and corporate training, so I care more about clear thinking than clever tricks. In my lessons we take one technique at a time, try it on...
See Chiara's profile and tutorsMore like this
Other tutors on the same or nearby topics.