Perfecting Our Modern Data Stack Recipe: What We Learned On The Journey At Huel
summary
Huel migrates data into Snowflake, uses Stitch then Fivetran for ingestion, splits ELT: Fivetran/Clean for loads and dbt reduced table run times from hours to minutes, saving Snowflake costs; DataRobot runs churn and CLTV models and automated sales reconciliation; ThoughtSpot grants 95 users building liveboards in minutes, cuts analyst request volume, and raises data literacy by exposing tables and DBT metadata; overall the stack proves performant and scalable for marketing and international growth.
AI-written summary · about
chapters (10)
- 0:00Introduction to the talk
- 2:50About Huel and its mission
- 4:14Data challenges before modern stack
- 5:18Building the core data stack
- 7:04Transitioning to DBT and ELT
- 8:07Expanding with analytics and AI
- 9:11Current data stack overview
- 9:54Introducing ThoughtSpot
- 14:08Live demonstration of ThoughtSpot
- 23:37Key takeaways and closing
related talks
50mMasterclass: Building a data stack in 2021
23mData Team as an Optimisation Problem
01h 35mThe Modern Data Stack
01h 28mA Guide to the Modern Data Stack
26mPutting your data warehouse to work - Reverse ETL & Operational Analytics
27mRecipe for Building a Discoverable and Governed Data Platform.
01h 17mData Engineering in the AI Era: Building a Robust Foundation for AI Applications
01h 03mDiscovering the value of a cloud data platform