January 22, 2020 1:00 PM
Data Products are the scalable, reproducible outcomes of a data science workflow. Deploying data products takes different technology than deploying traditional software products, as data products are founded both on code and data. Versioning, integrating, delivering, and deploying products based on a constantly changing foundation like data requires that you have a way to version data sets, training artifacts, and collect user feedback data in order to automatically trigger machine learning updates in a fully adaptive environment.
Organizations need these tools and processes in order to deliver the value data science teams create to end-users. Join us for a detailed description of how to implement data products and realize ROI from your data science teams.
What a Data Product is
What Data Versioning is
How data versioning is a critical piece of most machine learning-driven data products
Why Molecula’s VDSs are an ideal choice for data versioning
What a computational task graph is
How to build computational task graphs to drive machine learning-driven data products
Why task graphs are often critical for working with larger teams when deploying data products
Why machine learning CI/CD is important for all data products