mlops wrapped 2021 ✨ look how broken it is!
think we're solving deployment challenges fast enough? ha -- it's apparently only getting worse
(i suspect this is because more companies with lower ML maturity are deploying models, which is awesome, but isn't factored into this story via this chart)
(i suspect this is because more companies with lower ML maturity are deploying models, which is awesome, but isn't factored into this story via this chart)
with the rise of Models In Production, data scientists are becoming ops people too...
and yep, the fragmented prod ML stack is a real issue...
full report here: info.algorithmia.com
full report here: info.algorithmia.com
finally, i want to draw attention to the fact that this Enterprise AI/ML report had _no_ mention of ML algorithms or modeling frameworks. hopefully it is clear enough now that in industry, data management challenges largely outweigh ML & DL-specific problems
Loading suggestions...