Why Kanban is a great way to handle AI projects?

So, you are setting up an AI project: you have your team, you have your task, maybe some funds.

But how do you organize yourself? Will you just randomly go with the flow, or will you use an established methodology?

Maybe you think that a "well established" flow is restricting to such creative endeavor... but then again I beg you to think deeper.

 If you think of the classical model like Waterfall, going that road can be pure suicide in an industry that has major break troughs multiple times a year. So yes keeping it Agile is definitely the answer. 

While doing Scrum may be a bit too much, with all the daily meetings, and tracking and reporting, not measuring progress at all is a sure direction to failure ("What doesn't get measured doesn't get accomplished!" they say)

So a lightweight and flexible approach is an optimal in this case. This is where Kanban comes into play. 

Kanban is basically a board which has all the teams task, their priority(low/medium/high) and their status(ready/in progress/done). 

Obviously this can get more complex, even quite close to a Scrum standard approach, but its main benefit is precisely this: it is very adaptable to your project and to your team. There is no cookbook, no manual with rules. You make your own rules.

Personally I worked in both Scrum and Kanban style in an AI project. We started with Scrum but pretty soon discovered that meetings when needed are better than daily stand-ups, and that every team member can own their topic, and there is no need for a Scrum Master. Also Epics, and Stories and Tasks, with Parents and Children of X, was just over complicating our daily work and causing frustration.

Then when we switched to a simple Kanban board everything changed. With one look you could know who was working on what, what is the progress state, the deadline etc. Then you could freely ask questions or for support based on the others load.

Pretty relaxed, but at the same time having everything in order ;)