Championing modern science workflows to benefit dairy farmers
summary
DairyNZ replaces ad hoc files with a scalable stack: Snowflake cloud data warehouse, GitHub for code, and Posit Workbench for R plus Posit Connect to serve outputs. High‑frequency animal sensor data at ~60 observations/sec per device becomes ingestible, enabling machine learning projects and a ruggedized field app that streams to Snowflake. Training runs through the University of Waikato as spaced cohort courses with summative assessment, outputs include Shiny dashboards for farmers and a surprising finding that high profit, low emissions farms are geographically scattered.
AI-written summary · about
chapters (8)
- 0:00Modern science workflows intro
- 1:56Why change was needed
- 3:43Scalable data workflows
- 5:51Team skills development
- 8:43Business discipline & improvement
- 10:49Dairy farming use cases
- 15:46Key lessons & takeaways
- 17:16Q&A
related talks
20mData Science in Production: The Way to a Centralized Infrastructure
28mUsing Dataops To Build And Manage Data Products On Snowflake
25mPerfecting Our Modern Data Stack Recipe: What We Learned On The Journey At Huel
28mHerding Data to the Cloud
32mScalable Python and SQL Data Engineering without Migraines
46mThe challenges of data science for veteran media organizations
41mOn Data: Sky above, sand below, peace within
01h 17mData Engineering in the AI Era: Building a Robust Foundation for AI Applications