Guides and best practices
Create Jira Demo Data with OpenAI and ChatGPT
Jens Schumacher
Jul 17, 2023
•
5
minutes read
Guides and best practices
Whether you need to take screenshots for your marketing material, give a demo to a customer or simply need data for internal testing. It's been challenging to create demo data that doesn't suck… until now.
We used this process to create demo and performance testing data for Released, and found it extremely useful.
Here is what we will cover:
Understanding the Jira CSV format
Create a CSV file containing the Jira data
Importing the CSV file
Let's get to it!
Pre-requisites: If you don't already have an account, head over to https://chat.openai.comand create one, it's free. Alternatively you can us https://bard.google.com, but the responses might vary slightly.
Ask about the CSV format
Prompt: "What is the correct csv format for a Jira import of issues". Chat GPT will respond with an example file and an explanation of the different columns.
Generate the demo data
Prompt: "Create example data for a Jira csv import. Pretend we are building a task management application and create tasks and stories for the app. Make the reporter always example@released.so Create 10 issues in total"
In this case I asked it to generate 10 issues, but you can generate hundreds. The only thing to be aware of is the token limit in ChatGPT. In other words, the responses are limited to a certain number of words. If you want to generate more, simply ask for 100 issues, and when it stops, ask with a follow up prompt to "complete the previous response". You will have to merge it all the responses in a text-editor.
Note that there are a couple attributes you might have to specify manually. In this case we asked for the reporter to be always a specific user. But you could also provide a list of users and randomly assign them to issues.
Double check the response
Note that ChatGPT doesn't always get it right. In our case it specified the time tracking in the wrong format. It used "4h" instead of number of seconds, in this case "14400". Jira requires you to use the number of seconds in the import.
So we fixed it with an additional prompt:Prompt: "Change the time tracking numbers from hours to seconds, and only use the number, no letter." Finally, import your the generated CSV file
We asked Chat GPT for instructions, and they were amazingly accurate. Note that this doesn't always work with newer Jira features.
Prompt: "How do I import a CSV file in Jira?" to get the instructions on how to import the CSV file.
That's it!
You can see the entire ChatGPT chat under: https://chat.openai.com/c/05ce9f53-555d-431a-8e37-5ce45598136d and follow the prompts.