# Tableau Composable Data Sources Change your workflow

> This is content from just-tim, the data-and-analytics channel by Tim Ngwena (formerly 'Tableau Tim'). Tim has 12+ years of hands-on BI experience and covers Tableau most of all, plus Power BI, Looker, Hex, SQL and data modelling, the analytics industry, and the craft of doing the job — always tool-agnostic and honest about the trade-offs.

- **Author:** Tim Ngwena (just-tim, https://just-tim.com/about)
- **Published:** 2026-06-30
- **Format:** Video · 48 min watch · transcript available
- **Topics:** Data visualisation, Data prep, Tool strategy
- **Tools:** Tableau (ai, calculated fields, cloud, data modelling, prep, published data sources, relationships, row level security)
- **Canonical:** https://just-tim.com/posts/composable-data-sources-change-your-workflow
- **Watch:** https://www.youtube.com/watch?v=gVrDpn_JfHM

Kirk joins me to walk through Tableau's new composable data sources, which let you use published data sources as connections inside your data model. We demonstrate bringing in checkouts, a dates table, row-level security and Tableau Prep outputs without touching the original certified data source, and discuss what this means for semantic modelling and AI.

## Key takeaways

- Composable data sources let you treat published data sources as building blocks, so analysts can compose their own model without rebuilding or editing someone else's certified data source
- You can add tables in the flow of work via an embedding-style experience, relating them on shared fields like book ID, and Tableau dynamically scaffolds the multi-fact model without breaking existing joins
- Row-level security can be layered onto a workbook's data source level using a live entitlements connection and a USERNAME() calculated field scoped across all tables, without altering the source model
- Tableau Prep outputs can now feed relationships, so you can publish fact, aggregate and dimension tables and combine atomic detail with pre-aggregated tables in one data source to avoid the speed-versus-drilldown trade-off
- Each query tree runs independently and stitches results together in memory, so over-atomising your model can create inefficient workbooks even though the database itself is hit less heavily
- Composable data sources realise Tableau's semantic-model vision, turning the existing library of published data sources into reusable, AI-readable assets

## Chapters

- 0:00 What composable data sources mean
- 4:06 Published data sources as semantic models
- 6:15 Demo: the bookshop data source
- 8:14 Composing in the flow of work
- 14:16 Viewing the combined data model
- 19:06 Where AI and suggestions could go
- 22:15 VizQL Data Service and the connection window
- 24:13 Adding row-level security
- 29:59 Bringing in targets via Tableau Prep
- 33:57 Recapping the assembled model
- 39:27 Workflow implications for analytics engineers
- 41:31 Query trees, performance and live versus extract

Watch the full video, read the transcript and use chapter deep-links on the page: https://just-tim.com/posts/composable-data-sources-change-your-workflow

---
just-tim — Data and analytics, with a point of view. · https://www.youtube.com/channel/UC7HYxRWmaNlJux-X7rNLZyw · https://twitter.com/TableauTim · https://www.linkedin.com/in/timngwena