Expert Power BI Visualization

Your specialized agent for designing effective Power BI visualizations — covering chart selection, DAX for visuals, conditional formatting, custom themes, and dashboard layout best practices.

When to Use This Agent

Choose Expert Power BI Visualization when:

• Designing Power BI reports and dashboards with effective visual storytelling
• Selecting the right chart types for specific data stories
• Implementing advanced DAX measures for visual calculations
• Creating custom themes, conditional formatting, and interactive features
• Building executive dashboards with drill-through and bookmarks

Consider alternatives when:

• You need Power BI performance optimization — use a Power BI Performance agent
• You need data modeling without visualization focus — use a data architect agent
• You need general data visualization (not Power BI) — use a data viz agent

Quick Start

Copy
# .claude/agents/power-bi-viz.yml
name: Expert Power BI Visualization
model: claude-sonnet
tools:
  - Read
  - Write
  - Edit
  - Bash
  - Glob
  - Grep
description: Power BI visualization specialist for chart design, DAX visuals, themes, and dashboard layout

Example invocation:

claude "Design a sales performance dashboard in Power BI — include KPI cards, a trend chart with YoY comparison, a geo map, and a product category breakdown with drill-through to detail pages"

Core Concepts

Chart Selection Guide

Data Story	Best Chart	Avoid
Trend over time	Line chart, area chart	Pie chart
Part-to-whole	Donut, treemap, stacked bar	3D charts
Comparison	Clustered bar, grouped column	Radar charts for > 5 items
Distribution	Histogram, box plot, scatter	Pie chart
Relationship	Scatter plot, bubble chart	Bar chart
Geographic	Map, filled map, shape map	Tables with coordinates
KPI/Status	Card, gauge, KPI visual	Complex charts

Dashboard Layout Pattern

┌─────────────────────────────────────────┐
│  Title Bar    │    Filter Bar / Slicers  │
├───────┬───────┬───────┬─────────────────┤
│ KPI 1 │ KPI 2 │ KPI 3 │     KPI 4      │
├───────┴───────┴───────┴─────────────────┤
│                                         │
│         Main Trend Chart (60%)          │
│                                         │
├─────────────────────┬───────────────────┤
│   Category Breakdown│   Geographic View │
│       (40%)         │      (40%)        │
└─────────────────────┴───────────────────┘

Configuration

Parameter	Description	Default
theme	Visual theme (default, dark, corporate, custom)	corporate
interactivity	Interaction level (basic, cross-filter, drill-through)	cross-filter
accessibility	WCAG compliance level (a, aa, aaa)	aa
mobile_layout	Include mobile-optimized layout	true
max_visuals	Maximum visuals per page	8

Best Practices

1. Follow the 5-second rule for dashboards. A viewer should understand the main message within 5 seconds. Lead with the most important KPI or trend at the top-left. Use progressive disclosure — summary on the main page, details via drill-through.

2. Use consistent color encoding across all visuals. If "Revenue" is blue on one chart, it should be blue on every chart. Create a custom theme JSON that defines your color palette and apply it across all reports for visual consistency.

3. Limit each page to 6-8 visuals maximum. More visuals mean more queries, slower loading, and cognitive overload. Group related metrics, use tooltip pages for secondary details, and create separate pages for different analytical perspectives.

4. Use conditional formatting to encode meaning. Color scales on tables, data bars on measures, and icon sets for status indicators convey information faster than reading numbers. Red/amber/green for KPIs against targets is universally understood.

5. Design for the narrowest screen your users will use. If executives view dashboards on iPads, design for that viewport. Use the mobile layout editor in Power BI Desktop to create responsive layouts that work across devices.

Common Issues

Visuals are cluttered and hard to read. Too many data points, labels, and legends create noise. Remove gridlines, reduce label frequency, abbreviate axis labels (1M instead of 1,000,000), and use tooltips for detail instead of cramming information onto the chart.

Cross-filtering between visuals produces confusing results. By default, clicking one visual filters all others. This can produce misleading combinations. Use "Edit interactions" to explicitly control which visuals filter which, and disable cross-filtering where it creates confusing context.

Custom visuals are slow or don't print correctly. Custom marketplace visuals render differently than native visuals and often have performance issues. Prefer native Power BI visuals. When custom visuals are necessary, test performance with your actual data volume and verify PDF/print rendering.