# June 18, 2025 – Snowflake Semantic Models, LLM Search, Tableau Metric Extraction, and more

Leveling up your data workflows—this month’s update introduces automated semantic models, AI-powered search, Tableau Metrics generation, and deeper support for dbt and PowerBI.

## **✨ What's New**

### 🧠 **Auto-Generate** Snowflake Semantic Models **from BI Tools**

Most data teams struggle to build a semantic layer from scratch. With Select Star, you can now **auto-generate a Snowflake Semantic View, or Cortex Semantic Model YAML** directly from your existing BI dashboards.

We reverse-engineer the dashboards you already use by extracting or generating:

* **Column descriptions**
* **Primary/foreign keys**
* **Sample values**
* **Join relationships**

All pre-filled in a Cortex-ready YAML export or Semantic View to jumpstart your modeling.

<figure><img src="https://3470314135-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F-MgAiVthA_yg9UXKuhyY%2Fuploads%2Fgit-blob-3db5fbd42e041a595f0c6cc9a1f15d75f1682cfa%2Fimage.png?alt=media" alt=""><figcaption><p>Snowflake Cortex Model Generation</p></figcaption></figure>

Once your models are generated, you can start querying them directly in the Select Star UI.

<figure><img src="https://3470314135-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F-MgAiVthA_yg9UXKuhyY%2Fuploads%2Fgit-blob-c1e07b8a2bdfebf2b7977be746a2a2ae6db3c1be%2Fimage.png?alt=media" alt=""><figcaption><p>Cortex Analyst in Ask AI</p></figcaption></figure>

Learn more how to set it up in our [Cortex Analyst docs.](https://docs.selectstar.com/integrations/snowflake/cortex-analyst)

### 🔍 **LLM Search — Natural language search for your data catalog**

You can now use natural language to search across your data catalog. Instead of typing keywords, simply ask:

*“Where can I find customer churn metrics?”*

*“What dashboards use ARR?”*

LLM Search works alongside our keyword-based engine to help you:

* Understand user intent and surface more relevant results
* See AI-powered answers next to traditional search results
* Quickly navigate to datasets, dashboards, or documentation without exact matches

<figure><img src="https://3470314135-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F-MgAiVthA_yg9UXKuhyY%2Fuploads%2Fgit-blob-ce770f9831063b7ad64977a0221b0ed323619aac%2Fimage%20(169).png?alt=media" alt=""><figcaption><p>LLM Search</p></figcaption></figure>

Note: This feature is available only to customers who have **AI features enabled** for their organization. Learn more in our [Search docs](https://docs.selectstar.com/features/search).

## **➕** Improvements

### **🧠 Automated Metric Generation from Tableau (Beta)**

Following the PowerBI beta in April, Select Star now supports **automated metric extraction for Tableau dashboards**. Metrics are auto-documented with definitions, formulas, and context—saving hours of manual work while improving consistency and discoverability across your team. For more, check out the [docs](https://docs.selectstar.com/features/documents/metrics/metrics-generation).

<figure><img src="https://3470314135-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F-MgAiVthA_yg9UXKuhyY%2Fuploads%2Fgit-blob-36cdd94e6f6b6adb9244975a45c717ab226c6b90%2Fimage%20(162).png?alt=media" alt=""><figcaption><p>Metric generated from Tableau</p></figcaption></figure>

### **🔐 PBAC Enforcement in Lineage**

Your data access policies now apply consistently across all lineage views—ensuring users only see what they’re allowed to: Explore & Tree views, Source Tables and Columns, Downstream Dashboards, and CSV Export. This update strengthens data governance while keeping the user experience seamless. Learn more about PBAC in the [docs](https://docs.selectstar.com/user-management/policy-based-access-control).

### **✅** dbt custom test support

You can now see custom dbt tests in Select Star alongside built-in ones. This means:

* Custom tests are shown under the **Tests** tab and **Data Quality** sections
* Test details include the original SQL and any related models or tables
* If a test targets a single column, it’s shown; otherwise, it’s marked as table-level

<figure><img src="https://3470314135-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F-MgAiVthA_yg9UXKuhyY%2Fuploads%2Fgit-blob-900a3504db4c9197b7a2b4cbe3d2838885a466f0%2Fimage.png?alt=media" alt=""><figcaption><p>dbt custom test</p></figcaption></figure>

This makes it easier to trace and understand issues flagged by custom dbt tests. Especially for teams using validations for freshness, completeness, or business logic checks. No more digging through dbt code when something fails.

## **🔧** Incremental

**UX Improvements**

* **New Card View for Search Results:** Search results now support a cleaner, card-style layout — easier to scan and more friendly for business users.

**Integration Improvements**

* **PowerBI DirectQuery Support:** Power BI ingestion now supports **DirectQuery** and includes **upstream datasets**, giving you better visibility into data flow and dependencies.


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://docs.selectstar.com/changelog/june-18-2025-auto-cortex-models-llm-search-tableau-metric-extraction-and-more.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.