Leveraging AI in Project Management: Practical Tips

Generative AI took the world by storm. It made a difference in many fields, and it could certainly help in project management work in ways that go beyond summaries and reports.

Using AI in project management — photo by @steve on Pexels

CONTENTS ⇅

Free courses on using AI for project managers

Project Management Institute offers a free course on using AI, Practical Application of Generative AI for Project Managers. Another helpful free course by PMI is Talking to AI: Prompt Engineering for Project Managers. Passing them would yield PDUs and a Credly badge, and, of course, valuable insights on using AI in your project work.

This article started with the notes I took going through the Gen AI course, with more ideas and insights based on research and experience.

AI in project management: beyond the meeting summaries

The most common thing project managers mention when discussing AI usage is meeting minutes. AI takes the call transcript and transforms it into a concise summary with key topics and decisions made.

Not only does it save a lot of time, but AI analysis could also connect the ideas from different points in the conversation, so when you read back through the summary, you might notice the points that would give you further insights.

But summaries are only the beginning of what artificial intelligence could assist with.

Gen AI setup for a project management flow

What does using AI in project management actually look like?

Sometimes it would mean chatting with AI directly, but most likely the practical usage of Gen AI would mean setting up a system that pulls the data in, processes it with an AI system of your choice, and gives reports.

Such a system could be established using integration tools like Make or n8n. Make Academy has a free course on using AI agents, and the first couple of modules in particular give a very comprehensive breakdown of origins, strong points and limitations of Gen AI.

If you need assistance with system setup, I offer workflow setup and integration services.

Using Gen AI in Agile and Waterfall projects

Though “Generative AI” brings to mind the images of cutting-edge innovation and this evokes more associations with Agile approaches, it actually performs great when there’s a lot of past data and the current situation is similar to what there was before, and thus lends itself well to predictive approaches.

However, organizations used to classic approaches might have a harder time with adoption. According to the PMI report, most organizations that use Gen AI are using Hybrid approach to project management.

Generating traditional PM structures

Gen AI is useful in creating classic Waterfall structures, like Communication Management plan. Given the requirements, date constraints and personnel availability constraints, AI could produce Work Breakdown Structure, identify task dependencies, and perform cost calculation.

Uncovering patterns and predicting trends

Gen AI excels at using project-related information spread across various systems and documents to collect, extract and clean data. The data is analyzed to uncover and predict trends and patterns.

Risk calculation

One of the important aspects of trend prediction. In both Agile and Waterfall projects, AI could analyze historical data and monitor real-time data, like KPIs, to identify threats. Modern project management tools like Asana, ClickUp and Smartsheet have the AI risk analysis feature.

Market research and analysis

Gen AI could also monitor and analyze external data to stay on top of the industry news, see the market trends, regulatory changes, innovations, and competitor moves.

Creating internal and stakeholder reports

Reporting is another summary-type task that Gen AI is very good at. It would create all kinds of internal reports, as well as performance reports for stakeholders.

AI reports still need to be validated by humans, and AI needs to be regularly reviewed and trained to produce better output.

Generative AI as an expert opponent

On an individual level, a chat with AI could be turned into a discussion with a knowledgeable opponent who wouldn’t just give advice but would also challenge you, play the Devil’s advocate. Such a debate could lead to a fresh perspective and new ideas.

But to achieve this, you would need to create a persona to talk to, and shaping it would likely take more than a single interaction. But as a result you’ll get a worthy opponent willing to have constructive arguments with you at any time.

Considerations for using AI in project work

Insights are only as good as the data is

To properly calculate risk and other patterns, AI needs input that fully reflects the situation. So it’s important to keep all the statuses and due dates up-to-date, and put the tasks into the project management system to begin with.

I’m regularly asked which PM tool is “the best”. Depending on the specific flow, one tool might lend itself better over another, but ultimately the “best” project management system is the one everyone on the team is actually using.

AI and access to confidential data

When feeding data to AI, be careful with running confidential information through it. Implement the principles of data privacy, like anonymization and aggregation, access control, encryption.

The IT department has to approve the AI tools you’re using. If you don’t have a dedicated IT department, use your best judgement and keep the privacy concerns in mind.

Will AI take the jobs of project managers?

Throughout history, we’ve regularly seen fears of over-reliance on technology, with humans losing most jobs, or even the basic skills – it happened with the invention of the printing press, it happened with the calculator, and many times more. But human brain spent billions of years evolving to be as adaptable and versatile as it is, so it’s too early to write the humans off just yet.

Vijay Kanabar, director of project management programs at Boston University, underscores an AI education policy at his institution:

We do not want you to be editors. We want you to be the creators.

This approach encourages integrating GenAI into creative processes after initial human brainstorming to refine ideas effectively.

AI is not a replacement for but an enhancement for critical and creative thinking. Ultimately it’s the human who designs prompts and interprets data to make decisions. And human input and evaluation should be present at every key step of the way. The importance of “human in the loop” is highlighted in the PMI course as well.

What’s the best AI model to use in project management?

Among the two most popular LLMs on the market today, ChatGPT by OpenAI and Claude by Anthropic, the latter is a better choice overall for project management. It has better reasoning and coding capabilities.

Claude also has RAG (Retrieval Augmented Generation) for projects, which expands the capacity for referenceable project knowledge.

But there are other models as well, which might be more suitable for your specific needs, and different models might need to be used for different tasks. So first, decide what you want the model to handle, then look into the available options to see which one fits best.