Day 1: Choosing Where AI Adds Real Value

Selecting AI Use Cases

• Business problems with measurable value
• Assess AI suitability for problems
• Availability of required data
• Rank cases by feasibility

Measuring Data Readiness

• Availability of required data
• Assess current quality and completeness
• Required access and privacy conditions
• Identify major preparation gaps

Designing Analytical Prompts

• Define the analytical task precisely
• Provide necessary data context
• Specify the required output format
• Add essential verification constraints

Day 2: Using AI to Accelerate Analysis

AI-Assisted Data Cleaning

• Recurring data quality errors
• Suggest suitable cleaning rules
• Review changes before application
• Record of human-approved cleaning decisions

Generating Analytical Insights

• Request patterns that remain verifiable
• Link insights with numerical evidence
• Reject unsupported generated explanations
• Convert insights into follow-up questions

Detecting Anomalous Patterns

• Normal behavior baseline and thresholds
• Set suitable anomaly thresholds
• Review flagged cases manually
• Distinguish errors from real events

Day 3: Keeping AI Outputs Trustworthy and Responsible

Validating AI Outputs

• Outputs with original sources
• Recalculate important results independently
• Assumptions and possible hallucinations
• Document the human review outcome

Governing Responsible AI Use

• Clarify the use case objective
• Provide appropriate human oversight
• Record of consequential automated decisions
• Review ethical risks periodically

Integrating AI into Operations

• Require appropriate human review
• Design a review pathway
• Link outputs with decisions
• Measure continuing integration impact

• Analyze a professional case highlighting challenges in Selecting AI Use Cases
• Practical lab creating a workplace tool for AI-Assisted Data Cleaning
• Use a group exercise to test decisions around Validating AI Outputs
• Simulate a situation requiring Integrating AI into Operations in practice