Up to 85% of data-related projects fail. Most of that is due to ill management of these projects, and no one-size-fits-all solution (not even agile), can solve the problem. I want to share what I learned so far: from experience but also from good books and articles, start and promote a wider discussion on the data-related project and people management, and how we can help ourselves.
Depending on estimates between 60% and 85% of Big Data projects fail; we see similar numbers for AI/ML projects. Why is that? Have you ever been on a good- looking project that didn't deliver or was abruptly closed? A lot of that has to do with management, and if you think product based on software development was hard, data-related projects are even harder. In this talk I will combine my experience of being a part of or running teams in a few data-related projects, spanning from big data engineering to machine learning engineering in organization of a different size. I will tell you about my successes and failures, how I feel the data-related projects are perceived by the business and what can we do about this. The ideas and experience are backed by a number of books, articles and methods circulating around in the community. You will learn not only what you can do as a manager, but also as a team member. Also, that Agile is good, but is also not one-size-fits-all solution, and what is better to pass. By no means I feel as a know-it-all person in management, I just want to share what I learned so far, start and promote a wider discussion on the data-related project and people management, and how we can help ourselves.