With the advent of generative AI, software companies can now layer additional automation onto their existing processes. For example, software engineers can rely on generative AI tools to quickly produce and evaluate code. But the potential benefits of AI extend well beyond software development to other areas of tech, including project management, Agile, and Scrum.
Before we dig into how AI might impact project management and Scrum, let’s cover two quick points:
- First, don’t fear that these tools are going to “take your job.” Agile software development teams still need all the people they’ve always needed, including Scrum masters, product owners, and so on. These tools don’t replace the work; instead, they make your work more efficient.
- Second, although today’s AI tools seem to know everything and seem able to do everything (paint pictures, write code), in reality, they’re repeating what they’ve learned by reading documents or inspecting images or reading existing computer code. Yes, they’re modifying it and creating seemingly new things, but what they’re doing is a far cry from true human creativity. Additionally, they can do what computers have always been good at, such as processing large amounts of data and providing you with statistics.
AI tools are still a long way from truly possessing the abilities of the human brain, which brings us back to the point that they’re not going to replace your job. And that means you’ll want to consider these tools as your personal helpers as you figure out how to make your workplace’s projects and processes more efficient.
What Can AI Tools Do for You?
We’ve taken time to look at many different AI tools that can assist Agile software development teams. All the big project management software names now have AI plugins, and there are also third-party tools that integrate with many of these big names. We’ve compiled a list of features we saw, but first, let’s start with how you can just use a regular AI tool like ChatGPT to boost your Scrum and project management abilities.
When a project is getting started, the various Scrum team members will decide on the scope of the application, including its list of features in the form of user stories, as well as come up with a description and a mission or vision statement. This is where the language ability of the AI chat tools can shine: you can feed in all the text from the stakeholders on what they want the project to accomplish, and then ask the AI tool to create a list of user stories.
You could paste all the information into ChatGPT’s prompt, for example, and then ask it to provide a specific list of user stories to be used in an agile process. This would be much faster than you having to manually build such a thing.
Or, if you prefer to do it manually, you could take what you’ve built and ask ChatGPT to clean it up and make sure it’s readable and useful as a list of user stories.
In either case, you’ll still want to go through it carefully to see if it’s correct and complete. (Remember, AI is your helper, and you’re still in charge.)
Then you could ask ChatGPT to think of everything you just gave it, and then provide a good description and mission statement for the project.
The key here is to think about language. After the AI tool knows about the user stories and helped create the description and mission statement, you could ask the AI to break up the user stories into individual tasks to be placed in the backlog.
This step will show why the AI is only an assistant and not replacing your job. The AI will likely do a decent job of coming up with the tasks; however, it will likely miss a lot, and some of the tasks will be either too granular or not granular enough. So, you’ll use the AI to provide you with a starting point. You will then refine what it gives you, and then start placing the tasks in the backlog.
The AI tool could also help you prioritize the tasks and determine so that you can start assigning them for the first sprint.
At this point, you could probably continue with a prompt tool such as ChatGPT, but you’ll probably find it will be limited on just what it can do to help; it’s sort of a “backseat driver” at this point, offering advice through a lot of words, but not really interacting with the other tools.
Instead, you’ll probably want to upgrade to tools that interact with your project management tools. For example, as you’re adding notes to cards in a Kanban board and moving them around, a good project management AI tool could be watching and gathering data on your interactions throughout the project.
You’ll find that pretty much all of the project management tools now have AI plugins, and in the next couple of years you’ll see more and more such tools appearing. Many of the tools we looked at have a lot in common; there are also third-party tools that work together with existing project management tools. So, let’s talk about some common ways these tools can help you. Our goal here is to give you a sampling of what’s out there so that you can start thinking about what features you’ll need when it’s time to look for your next AI-powered project management tool.
Ordering Tasks in a Kanban Board
The new AI tools can look at the tasks in the backlog and help determine which ones should happen in the next sprint. Of course, you will get the ultimate say; the AI tool is simply making suggestions and certainly doesn’t know as much about the project and job as you do. But it might spot certain things; for example, it might know that a task has been sitting in the backlog for some time, but several other tasks are relying on its completion and therefore it should happen sooner than later.
How it determines this depends on a lot of factors; if this tool helped you in previous projects, it will learn more about how your particular team works. But it also will likely come pre-programmed with information the company that made it fed it, or pre-trained it with. (For example, a well-written task for a certain software component might require a certain software library or class to be complete. The AI tool can spot such dependencies quickly.) Several tools provide this; we found that Wrike and Asana have some great features in this area.
Analysis on How Long Tasks Take, Followed by Predictions
We’ve all encountered the situation where a specific task that’s slated for a single sprint takes much longer than planned. The task then continues into the next sprint, causing delays in other tasks that were relying on this one to be complete. Developers get stressed; stakeholders get frustrated or even angry.
But when the tasks are clearly spelled out (with the help of AI!) with great detail on what is expected, the AI tool can analyze the task to determine how long it should take. This analysis can be based on multiple factors, including how long similar tasks in other projects took to complete, as well as who is working on the task. The tool can then suggest breaking the task into two or more smaller tasks. The smaller tasks could be divided among the developers to complete in a single sprint, or have them spread out over multiple sprints. Then when the next sprint begins, there are hopefully no surprises and everybody is happy.
For example, an AI tool called Motion provides such functionality.
Analyzing Which Developers have Which Strengths
Over time, an AI tool can analyze what happened with the sprints, and will start to make connections between certain types of work and the developers who work on them. The tools might notice that one developer always finishes one type of coding on time, but tends to stall when doing front-end CSS work. As the tool makes such connections, it will then start to be able to automatically assign certain developers to certain tasks, again making your job a bit easier.
We found that Hive’s AI tools can do such work for you.
Take Notes During Standups and Retros
Some of the tools have the ability to listen to what people are saying (such as via Zoom or in person through one computer’s microphone), transcribe everything, and then build a summary of notes. Taking notes and compiling a summary is often a job most people don’t like to do, but somebody on the team will be tasked with it. Not anymore! Now the AI tools can do it. And because today’s generative AI is so good at language processing, it can create summaries and bullet-point lists that are likely as good as most humans can do. And that will save a lot of time.
Several tools provide this functionality; two we looked are called Spinach.io and ClickUp.
Reports and Visuals for Stakeholders
Presenting reports and diagrams to the stakeholders really isn’t a lot of fun for some people. The stakeholders are busy and sometimes impatient, and spending all your time building (and presenting!) perfect reports can be stressful.
You want the reports to be concise, well-worded, and give as much information as possible in as little space and time as possible. This is where the AI tools will help. Those that come with project management software are very good at knowing exactly what information to compile, and will be able to present it in a way that’s well-formatted, with good visuals, and with language that’s easy to read and right to the point.
Without AI tools, creating these reports manually can take a very long time, and that’s rarely much fun. But the AI tools can produce them nearly instantly. Then you can go through them, tweak as needed, and present them. That saves a huge amount of time, allowing you to focus on more creative tasks and interacting with the developers. (And maybe finally getting a 15 minute break to relax with a coffee!)
Monday and the aforementioned Wrike are examples of platforms that provide this feature.
A Few More Odds and Ends
In looking at the various tools, we also saw that some of them claim to be good at risk assessment and mitigation. This is a pretty advanced feature, but considering where AI is, it’s not surprising that there are tools in this space.
Another feature we saw occasionally is help with KPIs (Key Performance Indicators). These are those things that large corporations absolutely love and individuals working for the corporation usually hate. And anyone who has had to deal with them knows how difficult and grueling they are to put together. So this is another cool feature we’re seeing in some AI tools, and will likely see more of in the next year or two. Hive provides this feature, as does a product called Taskade.
Another interesting feature we even came across is a chatbot whereby members of the team can ask questions about the project itself. Now in theory, a good Kanban board should be able to tell much of the story about the project, but there are certainly times when the team members have to call the product owner or Scrum master over with questions the board doesn’t address. Such a chatbot is an interesting proposition, and perhaps this will become a standard feature as well.
Conclusion
As mentioned earlier, every single one of the big players now has AI plugins. You’ve almost certainly been getting emails and reminders from these companies telling you to check out their respective AI offerings. Try them out; you might be pleased with your increase in productivity. And remember: all tools are assistants—it’ll be quite some time before you need to worry about them replacing your management skills and creativity.
