In 2016 my department, located within a highly traditional health insurance company, decided to try something very unconventional: An experiment of Agile scaling in a way no one had done before. We threw the work on a wall and let a tribe of 50ish people self-organize into teams. And we repeated this exercise every two days! The system revealed there were many agile 'best practices' that we were accustomed to, that were no longer useful in this environment.

For almost two years we continued to evolve, and it worked extremely well. The project continued until some key environmental factors changed, ultimately killing the experiment. This paper outlines what we tried, what we learned, and what I would recommend to others, should they be interested in doing something similar.