In this workshop, we discuss our “Post-Modern Data Stack”, that is, a deconstruction of the MLOps stack we previously shared with the community. In particular, we join the “modern data stack” (Snowflake + dbt) with “modern MLOps” practices, using Metaflow to bridge the gap between data, training, and inference in a pure serverless fashion.
As usual, we refuse to work in a toy stack, and with toy data: leveraging our huge data release from last year, we walk through a real-world recommendation pipeline, going from raw data to a live endpoint serving predictions.
You can download: Joining the modern data stack with the modern ML stack: https://github.com/jacopotagliabue/post-modern-stack
In this workshop, we discuss our “Post-Modern Data Stack”, that is, a deconstruction of the MLOps stack we previously shared with the community. In particular, we join the “modern data stack” (Snowflake + dbt) with “modern MLOps” practices, using Metaflow to bridge the gap between data, training, and inference in a pure serverless fashion.
As usual, we refuse to work in a toy stack, and with toy data: leveraging our huge data release from last year, we walk through a real-world recommendation pipeline, going from raw data to a live endpoint serving predictions.