Feature Request: Business-Friendly Names, Synonyms & Numeric Formatting for Fabric Data Agents

# Feature Request: Support Business-Friendly Column Names, Synonyms and Numeric Formatting in Microsoft Fabric Data Agents

## 1. Feature Title
**Enable Business-Friendly Output Column Names, Synonyms, and Configurable Number Formatting in Microsoft Fabric Data Agents**

## 2. Problem Statement
Currently, Microsoft Fabric Data Agents often display technical database table and column names in the generated response tables, even when business-friendly names and synonyms have already been configured in the semantic model and provided through **Prep for AI → AI Instructions**.

For example, the output may display:
- `DIM_TERRITORY[TERRITORY_NAME]` instead of `Territory`.
- `FACT_ACTIVITY_SALES[XX_BRAND]` instead of `Brand`.
- `DIM_CUSTOMER[CUSTOMER_NAME_ID]` instead of `HCP Name`.

This exposes underlying database naming conventions to business users and creates confusion, reducing the usability and business readiness of Data Agent responses.
Additionally, numeric values are sometimes returned with excessive decimal precision, such as `0.00000000000000`, instead of appropriately formatted values such as `0.00` or `0.0`.

## 3. Requested Enhancements
### A. Automatically Apply Business-Friendly Column Names
Provide native support in Fabric Data Agents to use business-friendly names or configured synonyms for all output column headers instead of physical database names.

Requirements:
- Respect column display names, descriptions, and synonyms configured in semantic models.
- Allow business-friendly names to be defined through Prep for AI instructions.
- Apply these names consistently to generated DAX results, response tables, charts, and summaries.
- Ensure technical database names are not exposed in user-facing responses unless explicitly requested.
- Support mappings across multiple semantic models and source tables.
- Preserve the original physical column names internally for query generation and execution.

**Example:**
| Technical Column Name | Expected Business-Friendly Name |
|---|---|
| `DIM_TERRITORY[TERRITORY_NAME]` | Territory |
| `FACT_ACTIVITY_SALES[XX_BRAND]` | Brand |
| `DIM_CUSTOMER[CUSTOMER_NAME_ID]` | HCP Name |

### B. Support Automatic Column Aliasing in Generated DAX
Currently, using DAX `SELECTCOLUMNS` to explicitly alias output columns with business-friendly names works as a workaround.

For example:
`SELECTCOLUMNS(Table, "Territory", DIM_TERRITORY[TERRITORY_NAME], "Brand", FACT_ACTIVITY_SALES[XX_BRAND])`
However, this approach currently requires explicit instructions and does not appear to be consistently applied automatically.

**Requested enhancement:**
Make business-friendly column aliasing through `SELECTCOLUMNS` or an equivalent mechanism natively supported and generally available in Fabric Data Agents.
The Data Agent should automatically apply the configured business-friendly column names to the final result set without requiring users to hardcode `SELECTCOLUMNS` instructions for every query.

### C. Configurable Numeric Formatting and Decimal Precision
Provide native support for controlling the display precision of numeric values returned by Data Agents.
Currently, values sometimes appear with excessive decimal precision, for example:
- Actual output: `0.00000000000000`
- Expected output: `0.00`

Requirements:
- Support configurable decimal precision (0, 1, 2, or more decimal places).
- Respect numeric formatting configured in semantic models and measures.
- Automatically apply appropriate formatting to percentages, ratios, growth rates, and other numeric metrics.
- Avoid unnecessary trailing zeros and excessive decimal precision.
- Preserve the underlying calculation accuracy and avoid rounding the actual data during query execution.

## 4. Expected Benefits
- Prevent exposure of technical database and source-system naming conventions.
- Improve business-user experience and readability.
- Ensure consistent naming across multiple Data Agents and semantic models.
- Reduce dependency on lengthy AI instructions and custom DAX workarounds.
- Improve the presentation of numeric values and business metrics.
- Make Fabric Data Agents more production-ready for enterprise reporting and analytics.

## 5. Business Impact / Priority
-This enhancement is particularly important for enterprise implementations where Fabric Data Agents are embedded within Power BI dashboards and consumed directly by business users.
-Business users should see meaningful business terminology rather than underlying database structures. Similarly, numeric results should follow standard business formatting conventions.

**Request:** Please consider providing native, generally available support for business-friendly output column names, automatic synonym resolution, DAX output aliasing, and configurable numeric formatting in Microsoft Fabric Data Agents.