Master Power BI, Tableau, dashboards, and data storytelling to transform insights into compelling visual narratives
Transforms complex data into compelling visual narratives using Power BI and Tableau. Master dashboard design, data storytelling, and executive communication to drive business decisions through effective visualization.
/plugin marketplace add pluginagentmarketplace/custom-plugin-data-analyst/plugin install data-analyst-roadmap@pluginagentmarketplace-data-analystsonnetThe Visualization Architect role teaches you that data means nothing if nobody understands it. While Foundations Specialists clean data, SQL Experts retrieve it, and Statistics Specialists analyze it, Visualization Architects translate these insights into beautiful, interactive visualizations that drive action. This role bridges the gap between technical analysis and business decision-making, teaching the principles of effective communication through visual design.
Why This Matters: The best analysis loses impact if it can't be understood by stakeholders. Executives make decisions based on dashboards. Visualization expertise makes you indispensable to any organization and dramatically increases your influence and compensation.
This learning journey transforms you from a numbers analyst to a strategic communicator who can:
Timeline: 10-14 weeks of focused learning | Skill Level: Intermediate Strategic Communicator
Pre-Attentive Processing (< 250ms):
Users grasp these visual properties almost instantly:
✓ Color: Highlight important values
✓ Size: Emphasize magnitude
✓ Position: Most important - use x/y axes
✓ Length: Bar charts
✓ Orientation: Slope
✓ Shape: Differentiate categories
Slow to Process (requires conscious attention):
✗ Color saturation
✗ Angle
✗ Area (pie charts)
✗ Volume
Implications:
- Use position and length for quantitative comparisons
- Use color to highlight, not to encode numbers
- Avoid 3D effects (hard to compare)
- Remove decorative elements (chartjunk)
Visual Hierarchy:
1. Most important → Largest, brightest, highest contrast
2. Important supporting → Medium emphasis
3. Context/labels → Lower emphasis
4. Unrelated → Minimal or remove
Message: "Show Composition"
├─ Parts of a Whole
├─ Bar stacked 100% (BEST for comparison)
├─ Pie chart (Common but limited to few categories)
├─ Stacked area (Good for time series composition)
└─ Treemap (Good for many categories, hierarchical)
Message: "Show Comparison"
├─ Compare Categories
├─ Bar chart (BEST horizontal bars for long labels)
├─ Column chart (Vertical bars, limited categories)
├─ Slope chart (Compare before/after)
├─ Bullet chart (Include target/goal)
└─ Lollipop chart (Modern alternative to bars)
Message: "Show Trend"
├─ Show Change Over Time
├─ Line chart (BEST multiple series)
├─ Area chart (Total volume + components)
├─ Slope chart (Begin vs end point)
└─ Ribbon chart (Ranking changes over time)
Message: "Show Relationship"
├─ Correlation/Distribution
├─ Scatter plot (X-Y relationship)
├─ Bubble chart (3 variables: X, Y, size)
├─ Matrix/Heatmap (Two categorical + one measure)
└─ Network diagram (Connections between entities)
Message: "Show Distribution"
├─ Spread of Values
├─ Histogram (BEST continuous numeric data)
├─ Box plot (Quartiles + outliers)
├─ Violin plot (Distribution shape)
├─ Density plot (Smooth distribution curve)
└─ Dot plot (Individual values)
Message: "Show Performance vs. Target"
├─ Gauge/Speedometer (Single KPI vs. target)
├─ Bullet chart (Multiple KPIs)
├─ Waterfall chart (Contribution to total)
└─ Status dashboard (Multiple metrics with status)
Color Choice Psychology:
Red: Danger, urgency, loss (use for alerts)
Green: Positive, growth, success
Blue: Trust, cool, stable (professional)
Yellow: Warning, caution (use sparingly)
Orange: Energy, warning
Purple: Luxury, premium
Gray: Neutral, inactive data
Color Blind Friendly Palette:
✓ Blue, Orange, Green, Red, Purple, Brown (distinct)
✓ Use color + shape + text for distinction
✗ Red-Green combinations alone
✗ Blue-Yellow combinations on some monitors
Effective Color Practices:
1. Use one color for categorical data (different shades)
2. Use color to highlight specifically (white/gray for other)
3. Sequential colors: Light to dark for increasing values
4. Diverging colors: Light-dark-light for scale around midpoint
5. Use maximum 5-7 colors (more creates confusion)
6. Test with colorblind simulator (accessible-colors.com)
Example Palettes:
Colorbrewer2.org - Built-in accessible palettes
Tableau built-in colors - Designed for clarity
Google Material Design - Modern, tested palettes
Mistake 1: Starting Y-Axis at Non-Zero
Before: [Chart showing $95M and $96M with huge visual difference]
After: [Chart starting at $0, showing true proportion]
Impact: Can exaggerate differences by 10-100x
When OK: If you note it and have good reason
Mistake 2: Pie Charts
Issues: - Humans judge area poorly
- Hard to compare similar slices
- 3D makes it worse
Solution: Use 100% stacked bar instead
Mistake 3: Dual Axes
Issues: - Can manipulate scale to show false relationship
- Confuses viewers
Solution: Use separate panels or one common axis
Mistake 4: Too Many Dimensions
Issue: - Impossible to follow
- Overloads cognitive load
Solution: Limit to 3-4 dimensions max
Use filters or small multiples for more
Mistake 5: Decorative Elements (Chartjunk)
Remove: - Unnecessary 3D
- Decorative graphics
- Redundant labels
Keep: - Clean, simple design
- Focus on data
Mistake 6: Unclear Title/Labels
Bad: "Sales"
Good: "Monthly Sales Revenue (2024) - Actual vs. Target"
Mistake 7: Inconsistent Color Schemes
Bad: Red for positive in one chart, negative in another
Good: Consistent color meanings across all dashboards
Mistake 8: Missing Context
Bad: Bar showing $5M revenue
Good: $5M revenue (↓15% vs. last year, 12% below target)
Executive Dashboard:
├─ 4-6 key metrics only
├─ Executive summary at top
├─ Focus on performance vs. targets
├─ Minimal detail (drill-down available)
├─ Traffic light status indicators
└─ 1-2 minutes to understand
Analyst Dashboard:
├─ Comprehensive but organized
├─ Filters and interactivity important
├─ Detailed explanations
├─ Multiple ways to slice data
├─ Technical explanations acceptable
└─ Self-service exploration focus
Public-Facing Dashboard:
├─ Simple, compelling design
├─ Beautiful, polished appearance
├─ Minimal jargon
├─ Mobile-friendly
├─ Clear call to action
└─ Professional branding
Mobile Dashboard:
├─ Vertical layout preferred
├─ Single metric per section
├─ Tap to drill-down instead of filters
├─ Larger touch targets (44x44px minimum)
├─ Test on actual devices
└─ Load time critical
Power BI Components:
Power BI Desktop (Development):
├─ Get Data: Connect to data sources
├─ Data Modeling: Relationships, transformations
├─ DAX: Data Analysis Expressions language
├─ Visualizations: 80+ chart types
└─ Reports: Multi-page dashboards
Power BI Service (Sharing):
├─ Cloud publishing
├─ Dashboard sharing
├─ Row-level security (RLS)
├─ Scheduled refreshes
└─ Mobile apps
Power Query (Data Preparation):
├─ Connect multiple sources
├─ Transform and clean data
├─ Combine/merge datasets
└─ Incremental refresh
Data Model:
├─ Fact tables (detailed transactions)
├─ Dimension tables (descriptions)
├─ Relationships (fact to dimensions)
└─ Aggregations (pre-calculated summaries)
Step 1: Get Data
File > Get Data > Select Source
Options: Excel, SQL, CSV, API, Web
Load data into Power BI Desktop
Step 2: Data Modeling
Ensure tables connected properly
Review relationships (Manage Relationships)
Hide unnecessary columns
Create calculated columns if needed
Step 3: Create Visualizations
Drag fields to visualization
Choose appropriate visual type
Format colors, labels, titles
Add filters for interactivity
Step 4: Design Dashboard
Organize related visuals together
Use consistent color scheme
Add text boxes for context
Enable cross-filtering
Step 5: Publish
Publish to Power BI Service
Set up row-level security
Share with appropriate users
Schedule data refresh
DAX is Power BI's formula language. Common formulas:
Aggregation Functions:
=SUM(Sales[Amount]) -- Sum all sales amounts
=AVERAGE(Sales[Amount]) -- Average sales
=COUNTROWS(Sales) -- Count of rows
=DISTINCTCOUNT(Sales[Product]) -- Count distinct products
Time Intelligence:
=CALCULATE(SUM(Sales[Amount]), -- Sales for same period last year
SAMEPERIODLASTYEAR(Calendar[Date]))
=TOTALYTD(SUM(Sales[Amount]), -- Year-to-date total
Calendar[Date])
=DATEDIFF(MIN(Calendar[Date]), -- Days of data
MAX(Calendar[Date]), DAY)
Conditional Logic:
=IF(SUM(Sales[Amount]) > 1000000, "High", "Low")
=SWITCH(Sales[Region],
"East", "Region 1",
"West", "Region 2",
"Other")
Advanced:
=CALCULATE(SUM(Sales[Amount]), -- Sales where region = "East"
Sales[Region] = "East")
=VAR BaseAmount = SUM(Sales[Amount]) -- Variables for complex formulas
RETURN BaseAmount * 1.1
Ranking:
=RANKX(ALL(Products), [Total Sales]) -- Rank products by sales
Slicers (Filters):
├─ Column slicer
├─ Dropdown slicer
├─ Between slicer (for ranges)
└─ Relative date slicer (Last 30 days)
Cross-Filtering:
├─ Click bar chart → filters other visuals
├─ Hold Ctrl to select multiple
├─ Edit filter behavior (visual level)
└─ Bookmark combinations
Drill-Through & Drill-Down:
├─ Drill-down: Hierarchy in visual
├─ Drill-through: Navigate to detail page
├─ Right-click on data point to drill
Buttons & Navigation:
├─ Action buttons (navigate, bookmark)
├─ Bookmark buttons (save view state)
├─ Back buttons
└─ URL buttons (external links)
Performance Optimization:
├─ Limit visuals per page (< 10)
├─ Use aggregated data (pre-calculated)
├─ Cache reports locally
├─ Use bookmarks instead of filters
└─ Test on slow connections
Power BI Service Sharing:
Share with Users:
├─ Workspace (shared editing)
├─ Share report (read-only)
├─ Share dashboard (specific KPIs)
└─ Row-Level Security (user-specific data)
Row-Level Security (RLS):
1. Create role with filters
Example: Region = [User Region]
2. Assign users to role
3. Test with role filtering
App Publishing:
├─ Curate workspace
├─ Create app from workspace
├─ Control user access
└─ Update independently from source report
Gateway Setup:
├─ On-premises data access
├─ Scheduled refresh
├─ Real-time data refresh
└─ Configure credentials
Mobile Optimization:
├─ Mobile layout (separate from desktop)
├─ Large touch targets
├─ Single metric per screen
├─ Test on actual devices
Tableau Architecture:
Tableau Desktop:
├─ Connect to data
├─ Create worksheets (individual charts)
├─ Build dashboards (multiple worksheets)
├─ Create stories (guided narratives)
└─ Publish to Server/Online
Tableau Server/Online:
├─ Centralized publishing
├─ User management
├─ Scheduled refreshes
├─ Content distribution
└─ Performance monitoring
Data Sources:
├─ Excel, CSV, Access
├─ SQL databases
├─ Salesforce, Google Analytics
├─ Spark, Hadoop
└─ Published data sources
Data Types:
├─ Dimension (categorical): Blue pills
├─ Measure (numerical): Green pills
├─ Discrete vs. Continuous (affects chart type)
└─ Attribute vs. Value
Step 1: Connect & Explore
Connect to Data > Select source
Explore dimensions and measures
Create initial worksheet
Step 2: Build Individual Charts
Drag dimension to Rows/Columns
Drag measure to Rows/Columns
Apply filters and sorting
Format colors and styling
Step 3: Create Dashboard
Dashboard > New
Add worksheets to dashboard
Arrange using tiling/floating
Add titles, text, images
Step 4: Add Interactivity
Filters (dimension/measure/date)
Parameters (user-controlled values)
Highlighting (select to highlight)
Drill-down (show detail on click)
Step 5: Create Story
Story > New
Add worksheets sequentially
Add narrative text
Create guided exploration path
Step 6: Format & Polish
Use consistent color palette
Align objects properly
Hide unnecessary elements
Test usability
Basic Calculations:
=SUM([Sales]) -- Total sales
=AVG([Sales]) -- Average sales
=RUNNING_SUM(SUM([Sales])) -- Cumulative total
String Functions:
=CONCAT([First], " ", [Last]) -- Combine strings
=UPPER([Region]) -- Uppercase
=LEFT([Name], 3) -- First 3 characters
=FIND("east", [Region]) -- Find position
Date Functions:
=TODAY() -- Current date
=DATEPART("year", [Date]) -- Extract year
=DATEDIFF("day", [Start], [End]) -- Days between
Conditional:
=IF([Region] = "East", "Region 1", "Other")
=CASE [Status]
WHEN "High" THEN 1
WHEN "Medium" THEN 2
ELSE 3
END
Aggregation:
=SUM(IF([Profit] > 0, [Profit], 0)) -- Sum positive profit only
=COUNTD([Customer]) -- Count distinct customers
Window Functions:
=WINDOW_SUM(SUM([Sales])) -- Running sum across partition
=RANK(SUM([Sales])) -- Rank within partition
=PERCENT_OF_TOTAL(SUM([Sales])) -- % of total
Filter Actions:
├─ Click chart element → Filter other sheets
├─ Select state/region → Show relevant employees
└─ Custom field values → Filter by selection
Highlight Actions:
├─ Hover over bar → Highlight related items
├─ Color-code related data
└─ Multi-select using Ctrl
URL Actions:
├─ Click on cell → Open external website
├─ Use field values in URL
├─ Example: Open customer detail page
Parameter Actions:
├─ Create parameter (variable value)
├─ User changes parameter value
├─ Affects calculations/filters
├─ Example: Toggle between metrics
Go-to-Sheet Actions:
├─ Click element → Navigate to related sheet
├─ Pass field values to target sheet
├─ Create drill-down experiences
Tableau Server:
├─ On-premises installation
├─ User management
├─ Extract refresh schedules
├─ Row-level security
└─ Governance and audit
Tableau Online:
├─ Cloud-hosted by Tableau
├─ Automatic updates
├─ Easy collaboration
├─ Less IT overhead
└─ Higher cost
Tableau Public:
├─ Free sharing to public
├─ No password protection
├─ All data visible
├─ Use for portfolio/demonstrations
└─ Limited refresh options
Performance Optimization:
├─ Limit worksheets per dashboard
├─ Use published data sources
├─ Extract vs. live connections
├─ Pre-aggregate in database
├─ Eliminate unnecessary dimensions
└─ Test on various connections
The Narrative Arc:
1. Context (The "Why")
├─ Why does this analysis matter?
├─ What's the business situation?
├─ What question are we answering?
└─ What does the audience care about?
2. Conflict (The "What")
├─ What's the problem or opportunity?
├─ What challenges do we face?
├─ What gap are we addressing?
└─ Why should anyone pay attention?
3. Resolution (The "How")
├─ What does the data reveal?
├─ What's the key insight?
├─ What does it mean for the business?
└─ What's the specific finding?
4. Call to Action (The "So What?")
├─ What should we do?
├─ How should we act on this insight?
├─ What's the next step?
└─ What decision should we make?
Story Structure Examples:
Hero's Journey:
1. Hero (Our business) in ordinary world
2. Meets challenge (Problem in data)
3. Calls for action (Insight from analysis)
4. Transformation (New understanding)
5. Resolution (Recommended action)
Problem → Solution:
1. Here's the problem
2. Here's the data proving it
3. Here's what it means
4. Here's what we should do
Compare → Contrast:
1. Current state (before)
2. Data showing contrast
3. Desired state (after)
4. Path to get there
Annotation & Focus:
Technique 1: Highlight Key Values
Before: [Crowded chart with 20 data points]
After: [Chart with 1-2 values highlighted, others grayed]
Impact: Viewer focuses on your point
Technique 2: Annotate Key Insights
Use text boxes:
├─ Arrows pointing to key values
├─ Callout boxes with insights
├─ Trend annotations ("↑20% last month")
└─ Comparative notes ("2x industry average")
Technique 3: Color for Emphasis
Good: Red for problems, green for success
Bad: Multiple colors reducing focus
Technique 4: Small Multiples (Faceting)
Use: Show same metric across categories
Example: Sales trend for each product separately
Benefit: Compare patterns while keeping focus
Technique 5: Progressive Reveal
Build story piece-by-piece:
1. Show context
2. Reveal problem
3. Show supporting data
4. Highlight solution
5. Call to action
Avoid: Showing all at once (overwhelming)
Executive Dashboard Requirements:
Attention Span: 5-10 minutes max
Format: Single page (above the fold)
Essential Elements:
✓ Key metric (big number at top)
✓ Performance vs. target (red/yellow/green)
✓ Trend (is it improving?)
✓ Context (where does this rank?)
✓ Action item (what to do)
Layout:
╔══════════════════════════════════╗
║ Title: Key Business Metric ║
║ Current: $5.2M ↑ 12% vs LY ║
║ Target: $4.8M ✓ 8% above ║
╠══════════════════════════════════╣
║ Trend Chart │ Breakdown ║
║ (Last 12m) │ (By segment) ║
╠══════════════════════════════════╣
║ Alert/Action: Review Q4 plan ║
╚══════════════════════════════════╝
Language:
✓ "Revenue beat target by $400K"
✓ Numbers with context ("↑25% YoY")
✗ "Revenue was $5.2M" (no context)
Metrics to Include:
├─ Current value (big, visible)
├─ Change (% vs last period)
├─ Target/goal (are we on pace?)
├─ Rank/comparison (vs peer/industry)
└─ Trend (direction and momentum)
Technical Audience (Data Team):
├─ Show methodology and assumptions
├─ Include statistical measures (p-values)
├─ Acknowledge limitations
├─ Discuss alternative interpretations
└─ Invite technical questions
Business Audience (Managers):
├─ Focus on business impact
├─ Show practical significance
├─ Provide clear recommendations
├─ Use familiar terminology
└─ Include ROI/financial impact
Executive Audience (C-Suite):
├─ Big number first, context fast
├─ Focus on strategic implications
├─ One page maximum
├─ Clear decision asked for
└─ Minutes-long explanation, not hours
Hostile/Skeptical Audience:
├─ Anticipate objections
├─ Show robust analysis
├─ Admit limitations upfront
├─ Have alternative explanations ready
├─ Focus on what analysis proves, not disproves
General Public:
├─ No jargon whatsoever
├─ Relatable examples
├─ Beautiful, professional design
├─ Focus on "so what?" not "here's how"
└─ Tell human story, not data story
Scenario: Build executive dashboard for regional sales performance.
Objectives:
Deliverables:
Skills Applied: Dashboard design, data modeling, DAX/calculations, audience design, communication
Scenario: Visualize multi-channel marketing campaign performance.
Objectives:
Deliverables:
Skills Applied: Multi-source integration, storytelling, executive communication, design
Scenario: Build visualizations showing customer health and churn risk.
Objectives:
Deliverables:
Skills Applied: Scoring design, visual hierarchy, actionable insight, business impact
Months 1-2: Basic Competency
├── Understand visualization principles
├── Create basic charts in Power BI/Tableau
├── Understand audience design
└── Build first simple dashboard
Months 3-4: Intermediate Competency
├── Master Power BI OR Tableau deeply
├── Create multi-sheet dashboards
├── Add interactivity and filtering
├── Design for different audiences
└── Understand data storytelling
Months 5-8: Advanced Competency
├── Master both Power BI and Tableau
├── Build complex data models
├── Create compelling data stories
├── Lead dashboard initiatives
├── Design for scale and performance
Months 9-14: Expert Competency
├── Architect enterprise solutions
├── Lead visualization strategy
├── Mentor others on design/storytelling
├── Present to C-suite confidently
└── Ready for leadership roles
Entry Level (0-2 years): $65,000 - $90,000
Mid Level (2-5 years): $90,000 - $125,000
Advanced (5+ years): $125,000 - $160,000
Senior/Lead (8+ years): $160,000 - $220,000+
The Dashboard Grid:
3x3 Grid System (most common layout):
┌─────────────────────────────────┐
│ KPI 1 │ KPI 2 │ KPI 3 │
├──────────┼──────────┼───────────┤
│ │ │ │
│ Chart 1 │ Chart 2 │ Chart 3 │
│ │ │ │
├──────────┼──────────┼───────────┤
│ │ │ │
│ Chart 4 │ Chart 5 │Chart 6 │
│ │ │ │
└─────────────────────────────────┘
Design Rules:
1. Top: Most important metrics (KPIs)
2. Left-to-right: Importance decreases
3. White space: Prevents crowding
4. Alignment: Grid-based, clean
5. Hierarchy: Largest = most important
6. Navigation: Clear path for exploration
Dashboard Color Consistency:
Status Colors:
🟢 Green: Success, on-target, positive
🟡 Yellow: Caution, near-target, warning
🔴 Red: Alert, at-risk, problem
Business Metrics (Colorbrewer palette):
✓ Sequential (light→dark): Growth metrics
✓ Diverging (red←→green): Actual vs. target
✓ Categorical (distinct colors): Categories
Rules:
- Same metric = same color across dashboards
- Don't use red for positive even if company prefers
- Test with colorblind simulators
- Print in grayscale to check contrast
Dashboard Load Time Goals:
< 2 seconds: Excellent
2-5 seconds: Acceptable
5-10 seconds: Needs optimization
> 10 seconds: Critical issue
Optimization Techniques:
1. Data Level
├─ Query optimization (SQL)
├─ Pre-aggregation in database
├─ Incremental refresh
└─ Archive old data
2. Model Level
├─ Limit number of dimensions
├─ Hide unnecessary columns
├─ Optimize relationships
└─ Cache calculations
3. Visual Level
├─ Limit visuals per page (< 10)
├─ Simplify chart complexity
├─ Use appropriate aggregation
└─ Avoid continuous drill-down
4. Deployment Level
├─ Capacity planning
├─ Usage monitoring
├─ CDN for external content
└─ Regional servers if needed
Every Dashboard Needs:
1. Metadata
├─ Owner and contact
├─ Last updated date
├─ Refresh frequency
└─ Update history
2. Metric Definitions
├─ What does each KPI measure?
├─ How is it calculated?
├─ What data sources are used?
├─ Are there known limitations?
└─ What does green/red mean?
3. User Guide
├─ How to use the dashboard
├─ What questions it answers
├─ How to filter/drill-down
├─ Common questions (FAQ)
└─ Contact for help
4. Maintenance Plan
├─ Scheduled review (quarterly?)
├─ Metric updates needed?
├─ Performance monitoring
├─ User feedback incorporation
└─ Retirement plan if obsolete
Design for Everyone:
Color Blind:
✓ Use color + shape (not color alone)
✓ Test with Coblis simulator
✗ Red-green combinations
Vision Impaired:
✓ Large fonts (minimum 12pt)
✓ High contrast (70+ difference)
✓ Descriptive alt text
✗ Rely on color alone
Cognitive Overload:
✓ Maximum 4-6 charts per page
✓ Clear, simple titles
✓ Consistent layout
✗ Too many dimensions at once
Mobile Users:
✓ Responsive design
✓ Touch-friendly (44x44px minimum)
✓ Fast loading on mobile
✗ Hover-only interactions
Master design fundamentals
Set up tools
Build first dashboard
Master one tool deeply
Learn storytelling
Develop signature style
Prepare for advanced roles
Lead visualization initiatives
Current Role: Visualization Architect ✓ (You are here)
↓
Option A: Deepen visualization expertise
↓
Option B: Move to Phase 5 - Programming Expert
↓
Option C: Move to Phase 6 - Advanced Analytics
↓
Multiple Advanced Roles
↓
Career Leadership Roles (7 - Career Coach)
As a Visualization Architect, you'll understand that:
Your visualization skills can make mediocre analysis brilliant or brilliant analysis ignored. Master this craft.
Q: Should I learn Power BI or Tableau first? A: Tableau has gentler learning curve; Power BI integrates better with Microsoft. Learn whoever matches your company's tools. Both are valuable.
Q: How many charts should be on a dashboard? A: 4-6 for executive dashboards, up to 10-12 for detailed analytical. More than 15 and people stop looking.
Q: Why shouldn't I use pie charts? A: Humans judge areas poorly. A 100% stacked bar chart conveys the same information more accurately.
Q: What if stakeholders want 20 metrics on the dashboard? A: Educate them. A crowded dashboard is used less. Better to have 5 key metrics well-understood than 20 ignored.
Q: How do I handle conflicting stakeholder requests on design? A: Data and design principles should guide decisions. Show why certain designs work better, with examples.
Last Updated: November 2024 Difficulty Level: Intermediate Estimated Time to Completion: 10-14 weeks
Designs feature architectures by analyzing existing codebase patterns and conventions, then providing comprehensive implementation blueprints with specific files to create/modify, component designs, data flows, and build sequences