Agreed. It is as much a management challenge as an engineering one. Consolidating disparate data silos, iterating on the right algorithm architectures, setting

Agreed. It is as much a management challenge as an engineering one. Consolidating disparate data silos, iterating on the right algorithm architectures, setting up APIs in the right places, finding the right balance/split between AI and human efforts... You have to bring the right pieces together rather than just injecting the latest DL innovation into your enterprise (it won't work).