AI technology application for project management in Jira

Jira Development Jira Cloud
,
1

Hello!

I’m writing my bachelor’s degree work about AI technologies application for project management. For research part I need to explore how AI technologies help to complete project management tasks.

I have some questions and answers can help me with the research.
If some question is violating NDA, you may not answer to it.

  1. What kind of project management tasks are being solved using artificial intelligence technologies in Your product?

  2. Do You use any technologies to help project managers recruit staff/workers in a specific occupation for the project (for example, lead programmer)? If so, could You describe what parameters (for example, experience, interest for the job) the model takes into account?

  3. What kind of models are being used in Your product? (linear regression, neural network, solution tree)

  4. How did You gathered training data for the model? Did You use any publicly available sources of information?

  5. Does Your product has models trained without the “teacher/watcher”? If so, then why didn’t Your choice wasn’t made in favor of training with “teacher/watcher”?

  6. Which stage of creating the model do You consider as the hardest and most time consuming?

  7. What language/technology was used for model development? Can You recommend a technology for artificial intelligence model development?

Faithfully,
Diana

2

Your question is quite old, but lately AI is raising a lot of interest.

Even Atlassian has announced Intelligence:

We (as Marketplace Partners) are working on something similar to Atlassian’s product. Nowadays, with the OpenAI’s API’s is pretty easy to integrate AI as the OpenAI’s models understand SQL and are capable to write queries from Natural Language (NL).

We are going to perform a fine-tuning an Open AI model

For that, we have developed a SQL engine that runs on top of Jira’s Cloud REST APIs:

Vision SQL

which supports native JQL too.

Therefore, Jira users will write queries in NL. Then we will sent them to our Open AI fine-tuned model to get a precise SQL that we will run against Jira Cloud to resolve the data, Finally, we will transform such data into some table, chart, etc.

From the user’s point of view, they will write a NL query and will get a chart, table or something else automatically.

SQL looks one of the best approaches to integrate AI with SaaS and PaaS. ServiceNow acquired an AI company named Element AI and it seems they were working on using AI to generate SQL:

PICARD
DUROAT

We developed a prototype of Vision SQL for ServiceNow and in this video you can see what I mean about how is it possible transform NL into SQL (with OpenAI) and the run the query against a PaaS:

(Watch it as 2x speed)

As mentioned, it was a prototype hence some parts are not automated and are performed manually, but the idea is clear.

If you want to fine tune ServiceNow to use NL via SQL you can use this NPM package:

I think we will release our Jira Cloud integration with Open AI about September, 2023. As we support the full Atlassian’s REST APIs for Jira Cloud (Core, Software and Service Manager), I think Jira Cloud users will be able to ask anything in NL and get the data in a nice and visual format as well as interact with Jira.

Examples:

  • What boards can I see?
  • What open sprints have the board X?
  • What issues have the sprint Y?
  • Group they by priority
  • state
  • assignee
  • remaining time to due date order by less to more
  • change priority of issue Z from Critical to Major
  • resolve the issue W with this comment: “blah, blah, blah”
  • etc, etc, etc.
3

On a side topic, I know some ‘developers’ are already using ChatGPT to wrote or review their code for them.

This is from the Atlassian public forum:

Coding using ChatGPT
Coding using ChatGPT711×163 44.8 KB