Augmenting Human Decision Making with Data Science
summary
Claims that 95% of decisions happen in fast intuitive System One, that algorithms compute per-item match scores and an on‑the‑fly box likelihood, and that stylists retain veto power. Built systems filter inventory, show a top percentage of items, store feedback in a stats dashboard, escalate freeform complaints to support, and retrain models from overrides. Surprising numbers include radiologists contradicting themselves 20% of the time and a styling network of 3,300 stylists. Reported results: lower labor cost, higher keep rate, fewer mistakes, greater client satisfaction.
AI-written summary · about
chapters (17)
- 0:00Start
- 0:11AUGMENTING HUMAN DECISION MAKING WITH DATA SCIENCE
- 0:47I'm a software engineer at Stitch Fix - a personalized styling service
- 2:43How do humans make decisions?
- 3:19impressions & feelings
- 4:0895% of human decisions are made instantaneously
- 4:30analytical & effortful
- 6:41What are the limits in human decision making?
- 7:01Human decisions are unpredictable
- 8:12Human decisions are driven by your own past experiences
- 10:16What are examples of Stitch Fix stylists making decisions?
- 12:06In what ways can data science augment human decisions?
- 20:08Can humans augment algorithms?
- 22:19FOLLOW YOUR INTUITION
- 24:15What is the future of data science?
- 25:48predictive & intuitive
- 27:05What is the business impact of system 3?
related talks
34mAugmenting Human Decision Making with Data Science
18mSynthesizing human and machine capabilities
20mBrad Klingenberg, Director of Styling Algorithms, Stitch Fix
02mDBTB IN Eric Colson r
42mAn Algorithm of Style
32mLearning from algorithms to make Human decisions
36mHow can AI turn data overload into smarter decision-making?
34mSynergizing model and human decision-making using augmented machine learning