Managing AI Projects

Chapter 3. The Role of the AI Project Manager

The project manager is arguably one of the most underrated roles in the technology industry. Often seen as an overhead cost to be minimized, people in this role sometimes are made to work on too many projects at the same time and end up getting stretched too thin. Why spend valuable investors’ dollars on a role that doesn’t directly further the project’s progress? After all, PMs aren’t the ones training the models, are they?

Unlike traditional software development projects, AI projects are full of uncertainties. Their dependency on high-quality data and the need to constantly retrain the models make the AI project lifecycle look more like a plate of spaghetti than a beautiful linear Waterfall project or iterative Agile sprints. Unfortunately, the engineers and data scientists are often left to make sense of the mess on their own. That’s where the need for PMs comes in: to make the plate look less like spaghetti and more like neatly layered lasagna.

So the question is: What if the role of the AI project manager is not just helpful but actually critical to any AI initiative’s success? What if the many projects failing around the world could leverage advanced AI PMs with skill sets adapted to the complex nature of these projects to avoid continuous failures, cost overruns, and unfulfilled expectations?

In this chapter, we discuss the role of PMs in an AI project context, going well beyond the tasks of maintaining schedules and creating ...