DataOps 101 – A Better Way to Develop and Deliver Data Analytics
summary
Shows how to stop babysitting Power BI reports by adding concrete DataOps practices: put PBIX and model JSON into source control (git or OneDrive version history with ~500 versions), use PBI tools and Tabular Editor to diff changes, write automated tests with Pester/PowerShell or DAX Studio workflows, and wire Azure DevOps or deployment pipelines plus PowerShell/Power Automate to orchestrate CI/CD. Provides open‑source pipeline and monitoring templates (MIT), snapshotting/backups for reproducible data, and metrics to catch failed refreshes before users notice.
AI-written summary · about
chapters (12)
- 0:00Introduction to DataOps
- 3:39Common data analytics pain points
- 9:58Risks in data analytics projects
- 14:09DataOps Manifesto explained
- 18:05Five DataOps principles to apply
- 22:17Make it reproducible with versioning
- 24:46Quality is Paramount with testing
- 25:49Orchestration for automation
- 29:21Monitor quality and performance
- 32:56Reduce heroism to avoid burnout
- 37:50Practice what you preach
- 38:39Resources and closing remarks
related talks
48mGit off my lawn: Mastering Power BI Deployments with Git Integration
51mFrom DevOps for data to DataOps
20mDevDataOps - Using data for exponential growth of your organization
41mDataOps as a New Weapon for Data-Driven Companies
30mDevOps to DataOps
47mDataOps for AI-Ready Platforms
18mEmpowering Business Through Data
31mHow to Implement DataOps using