Podcast Episode 15: Project Management for AI Projects: Execution and monitoring

Successful management of AI projects

In today's fast-moving technology world, AI projects are no longer a thing of the future, but an essential part of modern companies. But how do you successfully manage such projects? In an interview with Julian Krätzmann, an expert in AI project management, we have gathered valuable insights and best practices for overcoming challenges in the execution, scaling and implementation of AI projects.

Key aspects for the implementation of AI projects

Julian emphasizes that the success of an AI project starts with the strategy. A clear AI strategy that is aligned with a company's digitalization strategy is essential.

This includes:

• Defining a vision and mission

• The development of specific use cases

• Setting clear goals

Another key point is data quality: “Clean and sufficient data volumes are the basis for every AI project.” Data protection, ethics and the consideration of risks also play a decisive role here.

Hybrid project management: a combination of agility and structure

AI projects require a combination of traditional and agile project management methods. Julian emphasizes that iterative development processes (e.g. through agile sprints) are particularly effective.

This should involve:

• Carried out regular tests,

• Check results and

• If necessary, data should be adjusted or supplemented.

Agile methods are supplemented by traditional processes, especially in the later implementation phases, to ensure a smooth transition to the production environment.

The importance of data and scaling

One of the biggest challenges in AI projects is to break down data silos and develop a centralized data strategy.

Companies must:

Julian describes an example from his practice: a forecasting tool for technicians that analyzes historical fault data. Here it became clear how important it is to integrate different companies in a network in order to improve the quantity and quality of data.

Key KPIs for AI projects

Julian distinguishes between three central KPI areas:

“An AI is only as intelligent as the data and feedback it receives,” explains Julian. Obtaining user feedback at an early stage is therefore crucial.

Promoting change management and acceptance

Successful implementation depends largely on employee acceptance. Julian recommends involving them in decision-making processes at an early stage.

Examples from practice:

• User involvement in the design of an AI system (e.g. when naming a phone bot),
• Creating an AI community to share knowledge and promote acceptance

A clear commitment to the topic of AI at management level is also essential.

The most important takeaways for AI projects

According to Julian, these are the decisive factors for the success of AI projects:

SUMMARY

AI projects are complex, but with the right strategy, a focus on data quality and an open corporate culture, they can be successfully implemented. For more insights and tips on AI projects, you can follow Julian Krätzmann on LinkedIn. He regularly presents tools and strategies there.