Jeff Jonas, IBM | IBM Edge 2013
summary
G2 is a platform that links diverse datasets to make high‑quality identity and behavior predictions; it is applied to voter rolls by joining state DMV records (including social security numbers) to remove duplicate registrations and help hundreds of thousands register and vote. Implementations target money‑laundering detection with a 180‑page technical document and pseudocode to feed G2, aim to lower false positives and false negatives, and are designed to run inside infosphere streams for low‑latency, context‑aware processing.
AI-written summary · about
chapters (10)
- 0:00Introduction and Iron Man stories
- 4:00Keynote on voter registration
- 5:45Introducing G2 technology
- 7:50How G2 finds data relationships
- 10:18Finding errors and lies in data
- 13:02G2 applied to voter registration
- 14:45G2 for money laundering detection
- 16:28Improving case quality with G2
- 18:11IBM’s big bets and future plans
- 20:40Enterprise Amnesia and flash storage
