Yes, I've written a data design sprint as this is what will need to be completed first to identify workflow and data owners. You can then run an Ai sprint once minimum viable data has been identified linked with user needs and business goals. I'm doing another sprint in the next month or so for UAE client.

@Christoph Piller the challenge all find with this is then IT. Enter legacy unstructured data systems, not standards, no data quality control. More often than not, I find this is the first failing point for AI. Even LLM OpenAI won’t save you. Bad data in = bad data out

AI Agents in workflows to support tasks between human and AI.