Data as Code with Dremio Arctic: ML Experimentation & Reproducibility on the Lakehouse

In this episode of Gnarly Data Waves, we will discuss how Dremio Arctic and data as code enable data science use cases like Machine learning experimentation and reproducibility on a consistent view of your data in a no-copy architecture.

As more data consumers require access to critical customer and operational data in the data lake, data teams need solutions that enable multiple users to leverage the same view of the data for a wide range of use cases without impacting each other. In this episode of Gnarly Data Waves, we will discuss how the data as code capabilities in Dremio Arctic enable data scientists to:

• • Create a data science branch of the production branch for experimentation without creating expensive data copies or impacting production workloads

• • Easily work and collaborate cross-functionally with other data consumers and line of business experts

• • Quickly reproduce models and results by returning to previous branch states with tags and commit history