Introduction
Data Studio AI Assistant enhances the Data Studio by helping users write SQL queries using natural language. This AI-powered assistant guides users through query creation, automatically validating and running queries to ensure accuracy. Users can interact conversationally with the assistant, refining queries and receiving immediate results, making it easier to generate insights, whether you're a SQL expert or a novice.
Effective prompting
Prompting an AI Assistant is a skill. We recommend reading the Prompt Engineering guide from OpenAI to learn more about effectively prompting an AI Assistant.
Some helpful tips for the Studio AI Assistant —
Write clear instructions.
It performs best when you give it clear instructions. Even little things like table names or explaining jargon really help.
Provide reference text.
Giving an example query and asking it to modify it in some way or providing examples of how you want it to respond can be very effective at producing desired results.
Use delimiters to clearly indicate parts of your input.
Using code fences around SQL can help the Assistant. Example -
Rewrite this query to show median lead time for all PRs.
```sql
SELECT open_to_merge FROM pull_requests
```
Clear instructions are the most effective way to get what you want from our AI Assistant.
Worse | Better |
Which team is fastest? | Show me the 5 teams with the lowest open to merge times for their pull requests. |
How long is it taking to finish work? | Build a query showing the average cycle time for completed issues in the past two weeks. |
Is the mobile team happy? | Whats the total number of deployments for the 'iOS' team last quarter? |