The modern data pattern

DOING ML EFFECTIVELY AT STARTUP SCALE

Replyable data processing and ingestion at scale with serverless, Snowflake and dbt

MLOps without too much Ops — Episode 4, feat. Jacopo Tagliabue, Ciro Greco and Andrea Polonioli

Introduction

As we discussed in the previous episodes, data is an integral part of ML pipelines: following the data-centric approach and then generalizing a bit, we may say that a sound pipeline for a data company is composed of two main blocks, dataOps (ingestion, …

data dataops dbt mlops mlops-without-much-ops snowflake