Is anyone working on automation of EA Practice activities such as design and/or roadmap development through the use of AI tools? This could include but not be limited to strategy, analysis, solution development, governance submissions, standards review, white papers etc. We are using ARIS as an EA repository and are considering the use of the AI companion or perhaps even standalone AI tools to import, process, analyses & publish. Happy to discuss ideas for future implementation.

3.2k viewscircle icon5 Upvotescircle icon2 Comments
Sort by:
IT Manager9 days ago

We've been working on an AI tool for a year now focused on artifact automated generation based on our standards and patterns. This tool started as a tool to help testers a QA Analysts to automate their job which was very hard to do. Now, this tool, which uses Google Gemini on the background, generates User Stories, Features, Epics, all integrated to our Azure Devops, creating dinamic documents with the use of GenAI to automate the hard work always considering complex software contexts. Our goal is to generate EA artifacts as well, along with, on the long term, generate code based on our EA documents and patterns. It's been a hard job to make it work altogether, considering AI halucinations, but it worths the effort because our users are already gaining more productivity and even thinking on things AI helps them to "think'".

VP of Engineering16 days ago

One area we have started looking at is to automate creation of software architecture/design documentation using some standard model like 4+1 view. We currently have documents created by different teams and we are looking to leverage a combination of copilots/agents to integrate them in better ways to tell an overall story. 4+1 view model does a good job of it, and we will try to automate creation of it using these tools. Of course, with oversight from architects and other technical leaders.

Content you might like

Ethical concerns18%

Cybersecurity/data privacy64%

Alignment with growth strategy36%

Change management 33%

Data or technology requirements10%

No concerns 2%

View Results

Poor quality code29%

Cybersecurity problems42%

IP/copyright concerns18%

Disrupted processes2%

Lack of attribution1%

Data leaks2%

Budget creep

Something else3%

No risks

View Results