Atlassian Intelligence: Meet your new virtual team member!

The buzzwords Artificial Intelligence (AI), OpenAI and ChatGPT have been buzzing around the media landscape for a good six months now. However, AI has been in use in numerous areas for much longer. Atlassian is now also relying on artificial intelligence and presented its version of AI (Atlassian Intelligence) at the Atlassian Team'23 event in Las Vegas.

Atlassian has been working with its own models for several years and in collaboration with OpenAI since the beginning of 2023. It was based on 20 years of data reflecting how millions of software, IT and business teams plan, track and deliver work and projects.

No user data was or is collected or tracked for this purpose, but rather movement patterns and team collaboration, e.g:

• Service-based work: teams taking inbound help requests and using custom workflows and data to find solutions for employees and customers.
• Project-based work: teams managing projects from concept to delivery with roadmaps, plans, tasks, goals and documentation.

The data protection principles and data protection guidelines were and are always applied.

This allowed Atlassian and OpenAI to analyze how and with which tools teams in a company work (together). Based on the analyzed information, a customer-specific teamwork graphic was developed and enriched with additional contextual information and data from third-party apps.

Atlassian lists the following main benefits:

1. Get instant help: Get instant help
2. Get up to speed: get up to speed with the latest developments
3. Accelerate work: Accelerate work
4. Learn on the fly: learn "on the side
5. Answer anything: answer everything

Here are some examples of the benefits:

1. Instant help: Get immediate help, for example via Slack Channel or Teams in the picture, the bot is called Charlie.
   
   
   
   
2. Get up to speed: e.g. summarizing the support feed and the activities of a helpdesk issue to bring the agent up to date.
   
   
   
   
3. Accelerate work: Confluence - for example, summarizing unformatted meeting notes and qualifying the information, e.g. recording decisions and derived activities, instead of having to manually edit such notes afterwards.
   
   
   
   
4. Leaning on the fly: Get immediate help at any point, e.g. technical terms on Confluence pages or texts in Jira issues.
   
   
   
   
5. Answer anything: Ask all kinds of questions
   
   • A: For example, general questions whose answers result from the combination and evaluation of various components (pages, blogs, etc.).
     
     
     
     
   • B: Creating a database query in natural language (Fig. 1), from which the corresponding query is created and executed in the query language (Fig. 2).
     
     
     
     

     Atlassian Intelligence creates the JQL query from this:
     
     

   • C: Query (in natural language) combined with the desired display mode
     
     
     

These are just a few examples and the models are constantly improving through continuous learning. Basically, what is suggested by AI is intended as support. These solutions can be adopted or adapted.

Further information on the topic of Atlassian Intelligence:

• https://www.atlassian.com/blog/announcements/unleashing-power-of-ai
  
• https://www.atlassian.com/software/artificial-intelligence
• Data protection guidelines: https://www.atlassian.com/trust/privacy/principles
• Privacy Policy: https://www.atlassian.com/legal/privacy-policy#what-this-policy-covers
  
  

Next Blog: News about Confluence