Using big data to inspire consumer confidence
summary
Credit Karma claims data on 20% of America's household debt, about 40% of U.S. Millennials and roughly one in three people who use credit, and it built a free tax product used by millions to join tax and credit signals. Engineering runs about a thousand model tests per month with models trained on ~120 billion observations, refreshed and applied in under 20 ms to drive recommendations, using strict probabilistic thresholds and a pledge never to sell user data to avoid harming consumers.
AI-written summary · about
chapters (9)
- 0:00Introduction to big data
- 1:47Credit Karma overview
- 3:40Building trust with data
- 6:27Overcoming consumer skepticism
- 9:18Data accuracy and financial stakes
- 11:44Real-time data processing
- 12:48AI and chatbots in FinTech
- 14:32Expansion and new markets
- 16:15Future of Credit Karma
related talks
31mData Science for Financial Inclusion
01h 19mOn financial modelling, credit scoring, and risk assessment
39mMachine Learning in Financial Credit Risk Assessment
28mBeyond the Balance Sheet: Transactional Data as a Window to Company Health
23mKeynote: Digital Psychometrics and it's Future Effects on Technology
50mData Science in a modern financial institution
22mThe Treasure Trove of Credit Data: Managing Risk
01h 13mGoverning credit scoring. Data protection, algorithms & surveillance.