Project management with AI: how the role of a project manager is changing in the era of generative artificial intelligence

Only 35% of projects launched worldwide achieve success. These figures were obtained by Standish Group based on a recent study. Jean Klushke, a project manager with a successful track record of implementing key projects in technology companies and startups, attributes the high failure rate to the lag in adopting AI and ML in project management. Jean served as our industry practitioner consultant for the preparation of this article. Together with the expert, we will examine the changes that the expansion of “artificial intelligence” into the profession brings to “project managers.”

Aymira Kuanysh, Editor at The Tech media

Jean Klushke is an experienced practitioner. His portfolio includes projects that brought millions of dollars to clients. Among them are Demo Days and Pitch Decks for investors, monthly follow-ups, and the systematization of project documentation. For some processes related to this work, generative AI was successfully applied.

Using new IT tools to improve project management efficiency can save businesses huge sums of money. There are figures showing that every year, investments in projects worldwide amount to $48 trillion. If 65% of these sums, according to Standish Group statistics, are simply wasted, then companies have strong reasons to quickly implement any tools that guarantee project results. AI can contribute precisely to improving effectiveness.

The future of project management: what will it look like?

According to the research agency Gartner, by 2030, 80% of tasks in the field of project management will be performed by AI. What will this look like? Imagine a specialist, for example, from a telecommunications company. The project status, progress toward achieving goals for seven key strategic projects of the company, he checks in a smartphone app. Two “taps” and the manager receives information about project charters, the level of involvement of key stakeholders, and even the morale of the team working on this project. Science fiction?

But the most popular platforms based on artificial intelligence, specifically ClickUp AI, Asana AI, Notion AI, Jadve, Trello with AI, already do much of what has been listed. The situation in project management is such that many AI tools for self-analysis, automated team management, and document processing have already appeared. Meanwhile, most project managers still rely on slides and spreadsheets, fearing new technologies or not knowing how to properly implement them in their work.

The problem with slides and spreadsheets is that they only work when project performance is measured solely by deadlines and results at a specific point in time. With modern adaptive work methods, such as Agile, such tools are poorly compatible. Among current project management tools, only AI-based portfolio management systems have gained widespread adoption.

AI tools in the field of project management today can:

— compare current project indicators with performance criteria established by the project manager and the team at the launch stage

— inform the CEO via mobile or web app about any changes requiring attention, as well as risks automatically detected by neural networks

— prioritize project activities by importance and offer solutions to the manager

— smart assistants, operating on a chatbot principle, will help during meetings: they will create a project plan and update it if decisions are made to change it. They will send this information to all team members and stakeholders. 

Under these conditions, the project manager will be able to focus on team training, developing support measures for projects and employees, and communicating with external stakeholders. If there is a drop in project performance, the AI tool will activate itself and recommend to interested team members to start a new “cycle” of the Agile process, so that the process of achieving project goals can be analyzed and reviewed faster, and then improved.

Five aspects of project management influenced by AI tools

Artificial intelligence increases the efficiency of project management, primarily through prioritization, more effective reporting to management on project status, documentation, testing, unique capabilities of generative AI, and “smart” project manager assistants.

1. Prioritization

Which project should be implemented first, and which later? Machine learning algorithms and neural networks have an advantage over human project managers in this question because they can quickly and comprehensively “process” all information related to the project. Additionally, artificial intelligence handles other tasks related to project prioritization better:

— better identifies projects for which the company already has much in place
— does not introduce human biases or perceptual errors into the analysis
— through comprehensive, rather than selective, information analysis, it better balances the project portfolio and risk assessment for the organization.

AI demonstrates its advantages only when it has access to high-quality datasets related to the project. If the data contains many errors, is unavailable, or unformatted, the quality of AI analyses and recommendations drops sharply.

2. Reporting to management

Project performance often suffers due to incomplete information received by management from departments responsible for PMO. With AI, it is much easier to establish effective circulation of information about progress in achieving project goals.

The most famous modern example is a project in the French public sector. Through it, the country’s president, Emmanuel Macron, can literally monitor metrics in key state projects in real time. Smart tools have radically transformed how PMOs work:

— monitoring based on performance criteria with AI is better
— AI can be entrusted with automatically generating reports and distributing them among team members; also, “smart” software can independently collect feedback on submitted reports
— more comprehensive information analysis with neural networks allows for more competent decisions on choosing project methodology
— monitoring compliance with company policies and procedures can be almost entirely delegated to the machine.

PMOs using AI can achieve an increase in efficiency by tens of times.

3. Project description, planning, reporting, risks

New AI-based applications for project management use Big Data analysis and machine learning for recommendations. This allows for more effective risk consideration, planning, and project description. Speaking of risks, AI systems can already provide recommendations for solving identified problems.

Soon, such software will be able to independently edit project plans and descriptions, modifying them to eliminate certain types of risks even at the work design stage. Improvements related to project description, planning, reporting, and risks in projects, driven by AI, are due to its better ability to:

— define the project scope, as it can take into account all user feedback without missing anything
— detail plans and allocate resources, as it “does not forget” any detail from the project documentation. 

In practice, preparing many types of project reports can be fully automated. For example, with proper AI implementation, the need to manually prepare monthly reports in projects disappears.

4. Virtual assistants in projects

Large language models, AI chatbots, and speech synthesis technologies have transformed many fields. Smart assistants are also appearing in project management. For example, Oracle recently released the AI assistant PMOtto.

If you give the bot a task: “Let the employee create a business card website by next week, allocate the necessary resources for this,” the bot’s response might look like this: “Based on the practice of previous projects of the selected employee, implementing the project you want to assign to them will take two weeks. Make corrections to the project plan?”

5. Advanced testing systems

At present, it is difficult to find a large company not using testing systems in development. With AI, their effectiveness and the range of tasks in the verification sphere that can be automated increase significantly.

The new role of the project manager

For many project managers, automation in the era of generative AI becomes one of the main tasks. True professionals are not afraid of “smart” tools but learn to apply them to gain advantages over competitors.

In the future, teams involved in project management may be viewed not as groups of specialists, but as a unity of human specialists and robotic systems. Specialists in this situation need to develop what AI cannot take on: soft skills, leadership, strategic thinking. Future project managers are responsible for strategy, leaving routine tasks to neural networks.

The importance of data preparation

We should also talk about data for machine learning. Any AI tool, smart assistants in particular, rely on pre-training AI on high-quality datasets. Without them, a neural network capable of generating solutions and offering recommendations lacks the context and “experience” to form any assessment of what is happening. Perhaps the organization launching the project has terabytes of data about previous projects or related to the project, but effectively training an AI assistant or other neural network software on them will not be possible. Perhaps this information is scattered across different files, presented in hundreds of formats, has different taxonomies, and contains many anomalies, i.e., unclear places that the neural network cannot process because the data is non-standard. Practice shows: approximately 80% of the time spent on AI implementation is used precisely for preparing data for machine learning. Datasets undergo many stages of processing before ML, including collection and cleaning. If you, as a project manager, have firmly decided to implement certain AI tools in your practice, you need to honestly answer a number of questions.

1. Are you ready to spend time, possibly months, reorganizing data for machine learning?
2. Are you ready to abandon prejudices or conservative habits in project management, such as monthly performance reports?
3. Will you allow the AI environment to make mistakes while the created software complex adapts to your context?
4. Are you ready to wait up to a year for AI tools to start truly “returning investments”?

If you answered “yes” to all questions, you are ready to apply project management tools based on artificial intelligence.

Project manager — robot?

Specialists in neural network technologies involved in creating automation tools for business processes in project management often predict the emergence of AI tools that will be able to manage projects independently, even without human involvement. But such predictions are made for many professions.

In reality, large language models, which have become the most popular tool in project management, are more like high-tech speech analyzers than artificial intelligence in the sense seen by science fiction writers.

This means that decisions still rest with humans, and the role of AI is currently that of a tool. A tool so important that the ability to work with it will determine who leaves and who remains in the market of professional project managers.