Day 1: Profiling Data Before AI Preparation

Structure, Completeness, and Consistency

• Validation checks for structure and completeness
• Business question tied to decisions
• Uncertainty limits around analytical conclusions
• Query logic checked against requirements

Missing and Invalid Values

• Performance signals around invalid values
• Consistent definitions for key measures
• Visual outputs informing decision clarity
• Explanation limits communicated to decision-makers

Duplicates and Anomalies

• Evidence requirements for duplicates and anomalies
• Validation checks before using results
• Outliers reviewed prior to model use
• Data quality before analytical interpretation

Day 2: Using AI to Improve Data Quality

Cleaning Rule Recommendation

• Control boundaries around cleaning rule recommendation
• Data lineage informing result traceability
• Assumptions tested before final conclusions
• Missing values requiring deliberate treatment

Entity and Format Standardization

• Review criteria for format standardization
• Prevention actions after recurring defects
• Acceptance evidence retained for oversight
• Control limits highlighting unusual variation

Anomaly and Error Detection

• Application limits around error detection
• Leading indicators of process drift
• Quality measures tied to outcomes
• Measurable acceptance criteria for quality

Day 3: Governing Prepared Analytical Data

Validation and Acceptance Rules

• Evidence requirements for acceptance rules
• Defect patterns needing root-cause analysis
• Failure patterns needing deeper investigation
• Corrective actions tied to evidence

Lineage and Change Records

• Application limits around change records
• Process variation before customer impact
• Verification checks prior to final release
• Sampling evidence supporting quality decisions

Data Quality Improvement Workflow

• Decision points for data quality
• Governance roles across the lifecycle
• Human oversight at critical decisions
• Escalation routes for failed controls

• Guided exercise applying invalid values to workplace cases
• Group review evaluating format standardization against evidence
• Case analysis examining acceptance rules in workplace decisions
• Simulation testing data quality under operational pressure