We're leveraging AI already in our openQM environment. We provide EDI managed services and EDI Integration services. To provide help on EDI elements and documents we are using AI to produce help screens. Rather than maintain a library of help within the application we make an API call to AI to provide summary or detail level help.

For our IT folks we have started experimenting with code documentation and release note generation. Surprisingly we can throw a CLASS written in OpenQM at AI and ask for description of what it does. AI returns a very accurate description of the class. We can then ask for a list of properties and methods in the class and it's very accurate. We are using this to build a code library as multi-value typically lacks intellisense in modern editors like VSCODE.  

Finally we leverage BitBucket for code version control, directly integrated into OpenQM. So we can modify code in VSCODE and push to an online repository. A webhook then delivers the modified code directly to our DEV environment. When we merge our develop branch to production branch a similar process occurs, pushing the code to production. When it's pushed to production we send the original code, and the new version to AI, asking for "release notes" and we get back very good release notes for what has changed between the two versions of the program(s). Our inside folks review the AI generated notes, make any adjustments if necessary and then send them out to our customers. The release notes are now a breeze to create and our system documentation is auto maintained.

Next experiment is to attempt to get AI to help with UML and class diagrams (as all our code in QM is object oriented).