Zenhub this week made available an edition of its project management platform for software engineering teams that makes use of generative artificial intelligence (AI) to summarize reports and identify bottlenecks to help streamline DevOps workflows.
Aaron Upright, the co-founder of Zenhub, said the company plans to work with each Zenhub AI customer to enable them to securely use their personal data to train an AI model using the ChatGPT platform to make DevOps teams more productive.
For example, DevOps teams will be able to use historical sprint data to identify bottlenecks more easily. Natural language processing (NLP) will also enable teams to reduce the impact of project label sprawl by making it easier to associate related projects regardless of naming conventions.
Overall, the quality of data being collected will also improve story point estimations, noted Upright.
It’s still early days as far as the usage of generative AI within the context of project management applications is concerned, but the potential AI has to transform the role of project managers is already apparent as more data is collected and analyzed by AI models. Rather than reacting to changes to project schedules as they occur, it should become easier for project managers to proactively surface issues that might have previously delayed or derailed a software development effort.
Project managers should also be able to better track the return on investment (ROI) metrics for any given project as changes occur, added Upright.
In theory, project managers augmented by AI models should be able to simultaneously manage more complex development efforts at levels of scale that previously would have been unachievable. The challenge is making sure project managers have visibility into the suggestions being made by AI platforms. Often when presented with conflicting data they will make up an inaccurate response, also known as a hallucination.
However, the narrower the range of data used to train an AI model the more accurate it tends to become. The critical issue many organizations will initially face is vetting the quality of the data being used to train the AI model. In addition, organizations will also need to make sure that no biases have been inadvertently injected into an AI model because the data used to train it was either flawed or inaccurate.
Ultimately, the amount of toil that project managers regularly experience should decline considerably as AI continues to advance. The expectation should be that AI will soon be pervasively applied across multiple project management applications. The race is now on to deliver on that promise.
In the meantime, project managers might want to start assessing which tasks they perform daily that are likely to be automated. This will enable them to add higher levels of value by having the time needed to focus more on strategic imperatives. After all, every minute spent checking to see whether data is accurate is one less minute spent understanding what needs to be done to affect a better project management outcome.