Analytics Dashboards
PSI’s quality analytics dashboards for tracking and exploring quality data.
Redbook Analysis Dashboard (v7)
Overview
The primary quality analytics dashboard, deployed to Azure App Service.
URL: https://ps-redbook-dashboard.azurewebsites.net
Authentication: Microsoft Entra ID (PSI credentials)
Tab Structure
| Tab | Purpose |
|---|---|
| Overview | Data exploration starting point |
| By Project | Project-level quality rankings |
| Root Causes | Pareto analysis, systemic issues |
| Trends | Time series, SLA compliance |
| Deep Dive | Preventability analysis |
| Products | Product-level repeat offenders |
| Explorer | Full data access with filters |
Tab 1: Overview
Purpose: Frame the data exploration journey, provide summary metrics.
Content:
- What’s in the data (projects, redbooks, date range)
- How we estimate cost (methodology explanation)
- When issues are found (detection timing)
- What’s causing issues (top root causes)
- Patterns across projects (systemic issues)
- Feedback form
Design Philosophy: Data exploration tool, not a report with conclusions.
Tab 2: By Project
Purpose: Rank projects by quality performance.
Content:
- Project Quality Ranking table (sortable)
- Scatter plot: Project Value vs Quality Cost/%
- New Build vs Retrofit comparison
- Quality by Industry and Process Type
Key Metrics:
- Quality % = Quality Cost / Project Value
- Target: <2%
Tab 3: Root Causes
Purpose: Identify what’s driving quality costs.
Content:
- Pareto Analysis with cumulative line
- Systemic Issues alert (>50% project spread)
- Prevention Point Mapping
- Earlier Detection Opportunity savings
- Who Opens Issues breakdown
Systemic Issue Definition: Root cause appearing on >50% of projects.
Tab 4: Trends
Purpose: Show how quality is changing over time.
Content:
- Monthly volume trend
- Monthly cost trend
- Cost per ticket trend
- Detection timing distribution
- SLA Compliance by Severity
SLA Targets:
| Severity | Target |
|---|---|
| Critical | ⇐3 days |
| Major | ⇐14 days |
| Moderate | ⇐21 days |
| Minor | ⇐30 days |
Tab 5: Deep Dive
Purpose: Analyze preventability of top issues.
Content:
- Guided narrative through findings
- Key finding: 94% preventable
- Two-column review interface
- Feedback forms for human validation
Workflow:
- View issue in left panel
- Review details in right panel
- Submit feedback/corrections
- Navigate to next issue
Tab 6: Products
Purpose: Identify product areas where issues cluster.
Content:
- Top Product Classes by cost (Pareto)
- Class x Root Cause Heatmap
- Multi-project product areas
- Quality Risk Scoring (0-100)
- Design reuse analysis
Important Caveat: Drawing field is a “pointer” to problem area, not precise fault location.
Tab 7: Explorer
Purpose: Drill down into specific records.
Content:
- Filter dropdowns (Category, Severity, Root Cause, Project)
- Text search (ID, Project, Problem text)
- Column group selection
- Row detail view
- CSV export
Jump-to Navigation: Buttons throughout dashboard link to filtered Explorer views.
Sidebar Controls
Filters
| Filter | Options |
|---|---|
| Ship Year | Multi-select by year |
| Industry | Multi-select by industry |
| Exclude Tasks | Toggle to hide ~716 task items |
Cost Model
| Setting | Default |
|---|---|
| Engineering Rate | $150/hr |
| Shop Rate | $125/hr |
| Stage Multipliers | 1.0x - 2.0x |
| Calibration Factors | Eng 1.0x, Shop 0.87x |
Admin Features
Access: Enter the admin password in Overview tab. Contact IT for credentials.
Unlocks
- Engineer Analysis tab (Quality % by engineer)
- Calibration controls
- Feedback analysis dashboard
Engineer Analysis
- Quality % by Mechanical/Controls/Proposal Engineer
- Mean vs Median toggle
- Project count in labels
- Top Project Concentration metric
Running Locally
# Activate environment
cd C:\Users\AMD\OneDrive - Progressive Surface\Documents\GitHub\Redbook
# Run dashboard
streamlit run Scripts/survey_app_v6.pyNote: Local runs use development user email from environment variable.
Legacy Dashboards
v6 Backup
streamlit run Scripts/survey_app_v6_backup.pyOriginal detailed dashboard preserved for reference.
v4.28 (Validator Mode)
streamlit run Scripts/survey_app_v4.28.pyLegacy dashboard with validation features for comparing AI vs human classification.
Related Pages
- Methodology - How metrics are calculated
- PSI Data Brain - Master data source map (start here for any data question)
- Data Dictionary - Field definitions
- Data Brain - Data sources
- Redbook Analysis - Pipeline details
Last updated: February 2026