Don’t tell anyone, but there’s a rumor going around that artificial intelligence practitioners are playing favorites. Seriously, their training models get all the love. Companies throw resources at developing robust training models, but the basic infrastructures surrounding them are often ignored to the extent that companies underestimate the level that performance plays across the board. Here’s what they do:

 

  1. Underestimate the need for scale when building their AI infrastructures
  2. Overfocus on model training at the expense of data cleansing and prep work
  3. Forget to remain agile as they leverage AI 

 

We are bringing together Scott Sinclair, who is an IT veteran, and Ken Grohe, the CRO of WekaIO, both of whom spend many hours talking with infrastructure managers and data scientists about the future of their data pipelines and putting their AI models into production.