Interview with Remus Lazar, Executive Director at IBM Watson, Moving Models From Experimentation to Production

For large organizations moving data science projects for experimentation to production have always been a challenge.  It requires careful collaboration with both data scientists, technologies and business end users. Once machine learning models are in production it is another challenge to keep them performing well.  Sophisticated workflows now provide monitoring features, like model degradation tracking and swapping out production models automatically without impacting endpoints is key.

View Remus’ ODSC West 2017 Keynote

View the same talk with american sign interpretation.