Are you really safe?
summary
Company AVA aggregates thousands of data sources into a single risk score that combines past, present and predictive models for locations worldwide. System claims detection of the 2017 London Bridge incident 4 minutes before traditional procedures and hourly crime forecasting with an error of ~5% (predicting 15 stabbings versus 14–16 observed). Tools include social media trust scoring, mobile cell network anomaly detection, radio dispatch transcription, ambient audio and gunshot triangulation tested in Barcelona to pinpoint incidents in real time.
AI-written summary · about
chapters (21)
- 0:00Introduction
- 4:48AVA Risk Score
- 6:34Context Sensitive
- 7:57Data
- 9:34Incident risk prediction
- 10:40Deep contextualization
- 11:50Digital media monitoring
- 15:20Trust scoring
- 16:12Personal data
- 17:26Listening to police
- 19:16Ambient audio detection
- 19:54Audio triangulation
- 21:06Cameras
- 23:16Repeat Patterns
- 24:22Additional Information
- 25:00Field Staff
- 25:21SDK
- 25:52Data influence
- 26:13Financial markets
- 26:24Hotels
- 26:53Thank you
related talks
16mDeep Tech From Berlin & Novi Sad
44mReal Time Vulnerability Alerting by Using Principles from the United States Tsunami Warning Center
21mEnhancing Threat Intelligence Using AI
27mAgentic Cyber Defense with External Threat Intelligence
47mReal Time Vulnerability Alerting
25mReal-Time Fraud Detection at Scale
17mUsing AI to anticipate risk
29mThresholds Are for Old Threats: Demystifying AI and Machine Learning to Enhance SOC Detection