AI for Data Analytics

Trust but Verify

• Why verification is taught first: AI failure modes including hallucinations, wrong methods, and context blindness
• The 7-step AI Validation Checklist for systematically evaluating any AI-generated analysis
• Live hallucination example: seeing how AI fabricates plausible statistics and fictional citations
• Introduction to the AI Traceability Document for professional accountability

The AI & Analytics Landscape

• The analytics maturity curve: descriptive, diagnostic, predictive, and prescriptive analytics
• AI taxonomy for analysts: how machine learning, deep learning, and generative AI relate to data work
• The ACHIEVE framework for deciding when AI adds value vs. when manual methods are better
• Bias and fairness in AI: real-world examples and how to incorporate fairness into your verification practice

GenAI as Your Analytics Co-Pilot

• The AI-augmented analytics workflow: Import, Clean, Explore, Analyze, Visualize, Report, Verify
• Hands-on lab: clean a messy dataset, generate statistics, ask analytical questions, visualize findings, and verify results
• Understanding the “dirty data” problem: how AI automates cleaning but requires your judgment on every decision
• Why “clean” doesn’t mean “perfect”: recognizing data quality issues that survive automated cleaning

Prompt Engineering for Data Work

• Three things every analytical prompt needs: role, task with data specifics, and output format
• Six prompting patterns for analysts: Describe, Explore, Compare, Predict, Explain, Validate
• Iterative prompting techniques: Refine, Redirect, Constrain, and Challenge
• Comparing AI tools: running the same prompt in different tools and evaluating where they agree and disagree
• Building a personal prompt library of tested, reusable prompts for real job tasks

Predictive Analytics Demystified

• Core concepts: regression, classification, and clustering — when to use each, no math required
• Key metrics: R-squared, p-values, accuracy, precision, recall, and the train/test split
• Hands-on lab: build a classification model, evaluate metrics, write data-backed recommendations, and self-critique
• Defending AI-assisted findings under stakeholder questioning using your traceability document

Critical Evaluation & Responsible AI

• Progressive verification: detecting Simpson’s Paradox, confounding variables, selection bias, and overfitting
• Finding subtle errors in professional-looking AI analyses through structured evaluation exercises
• Applying the full validation checklist collaboratively at speed
• Data privacy and governance: when NOT to upload data, and regulatory considerations (HIPAA, FERPA, GDPR, FISMA)

AI Tools, Chain Reaction & Live Problem-Solving

• The 2026 AI analytics tool landscape: ChatGPT, Claude, Copilot, Gemini, Tableau AI, and ThoughtSpot
• End-to-end automation demo: from raw data to stakeholder-ready executive brief in minutes
• Live problem-solving: a real work problem solved with AI in real time, unrehearsed
• Advanced techniques overview: NLP for text analysis and time series forecasting

Capstone

• Redesign a real workplace workflow with AI tools, verification steps, and traceability built in
• Map the before and after: current steps, tools, and time vs. the AI-augmented version
• Estimate time savings, identify risks, and define a concrete first implementation step
• Present and defend your redesign in a mini stakeholder simulation