Unleashing the Power of Analytics Engineering
summary
Analytics engineering converts a messy modern data stack into reliable, decision-ready reporting: missing/incorrect data largely disappear, refresh cadence moves from once-a-day to every two hours, analysts stop spending most of their time cleaning and instead deliver insights, and business users start acting on data. Implementation includes dbt testing and observability (Datafold/Monte Carlo), CI/CD with branch and PR schemas, version control, centralized dbt documentation exposed to ThoughtSpot, and they work.
AI-written summary · about
chapters (25)
- 0:00Analytics engineering revolution
- 1:04Origins of analytics engineering
- 2:07Early scalable data struggles
- 3:10Data reliability issues
- 5:58Manual process root causes
- 7:01Analytics engineering as librarian
- 10:10Building data function
- 11:54Measurable improvements
- 12:38Avoiding tech debt in stacks
- 13:20Learning tools and communities
- 14:22Naming conventions importance
- 15:26Prioritizing tool changes
- 16:10Automated testing for reliability
- 18:18Continuous integration and delivery for data pipelines
- 20:25Version control for data code
- 21:51Testing environments for safety
- 22:12Centralizing documentation
- 23:14Documenting data assets
- 24:40Documentation in workflows
- 26:08Close to business users
- 27:10Collaborating with stakeholders
- 28:35BI tool limitations
- 29:58Training analysts in engineering
- 31:01Key takeaways
- 31:46Q&A
related talks
30mData Engineering building your BI infrastructure from scratch
40mData Engineering in a Rapidly Changing World
03mDBTB INT Daria Mehra
38mThe Rise & Downfall of the Data Engineer REVISITED
25m"Data Engineering: The World Between Worlds"
50mDataOps 101 – A Better Way to Develop and Deliver Data Analytics
51mPanel Debate: Rise Of The Data Engineer
31mFuture of data engineering: what's in store?