Finding insights you can trust

AI for Data Analysts

Andy Cotgreave

Author and Speaker, Former Technical Evangelist at Tableau

What AI is doing when it "finds insights?"

under_the_hood.jpg

AI for Data Analysts

Four useful starting points

  • Trends over time

trend.png

  • Distributions across categories

distributions.png

  • Differences between groups

group_differences.png

  • Outliers and anomalies

outliers.png

AI for Data Analysts

Asking with intent

insight_prompts1.jpg

AI for Data Analysts

Asking with intent

insight_prompts2.jpg

AI for Data Analysts
AI for Data Analysts

Signal vs. noise

ice_cream_shark_attacks.jpg

AI for Data Analysts

Four checks before you trust an insight

 

✅ Calculation: does the query hold up?

✅ Mechanism: is there a plausible reason, or just coincidence?

✅ Stability: does it survive slicing (time, segments, cohorts)?

✅ Domain knowledge: does it match what you know about the business?

AI for Data Analysts

Let's practice!

AI for Data Analysts

Preparing Video For Download...