Visualize more, spend less: Unlock data visualizations with Apache Superset
summary
Demonstrates embedding Apache Superset into a SaaS product by connecting a database, creating physical and virtual datasets, and building a dashboard that reduces a 50,000‑row table to ~200 rows for fast charts. Shows chart types, image columns in tables, scheduled reports, SQL Lab tuning, and dashboard-as-code deployed from Git. Uses the Unembedded SDK plus a backend that issues short‑lived tokens, runs on Kubernetes with caching to cut query load, and supports a production single Superset instance serving 1,500+ users and reduced 50,000 rows to ~200 via virtual dataset while embedding via Unembedded SDK with short‑lived token.
AI-written summary · about
chapters (26)
- 0:00Introduction and mission
- 1:07Data as our superpower
- 2:30Apache Superset overview
- 4:15Adding a database connection
- 5:18Creating data sets
- 6:42Visualizing robot production
- 8:57Chart creation and customization
- 10:19Daily activity line chart
- 12:08Adding robot images
- 13:12Detail view per robot
- 14:15Building the dashboard
- 15:41Sharing and downloading
- 16:25Embedding dashboard in web app
- 18:09Security with tokens
- 19:33Demo success and features
- 20:38Miracle’s production use case
- 22:26Dashboard as code
- 23:13Q&A: Metabase comparison
- 24:16Q&A: Power BI vs Tableau
- 25:40Q&A: Data governance
- 26:00Q&A: Migration cost
- 27:03Q&A: Resource requirements
- 28:45Q&A: Dataset size limits
- 29:48Q&A: SQL debugging
- 30:30Q&A: Monitoring instance
- 31:11Q&A: Big data performance
related talks
34mQuantifying the growth and evolution of open-source communities with Apache Superset
41mEffective Dashboard Design Using
30mData @Scale Seattle- Pawel Terlecki
35mScaling Interactive and Insightful Dashboards with Data Studio and BigQuery
41mVisualizing big data
32mUsing Dashboards to Help Manage All Community Channels
53mVisualizing the Output of Your Search AI
58mPower BI for Developers