Fighting Financial Crimes as an SRE
summary
Concrete claim: SREs can stop fraud by wiring log collection, detection and automated response on AWS. The talk shows a working pipeline using CloudTrail + VPC flow logs and CloudWatch to spot assumed-role misuse and traffic floods, Macie for DLP to block sensitive-file writes, and Kinesis Firehose → S3 → Glue → EMR Spark plus Lambda to run scheduled analytics and alert. Cites real failures (3.5M fake accounts, €2B missing, 76M compromised) and demonstrates detection rules for assumed-role access, failed logins and sensitive-data modification.
AI-written summary · about
chapters (11)
- 0:00Fighting financial crimes as SRE
- 3:14Financial crime failures in banks
- 6:27SREs prevent financial crimes
- 9:16Secure cloud infrastructure on AWS
- 12:03Insider threats with CloudTrail
- 14:36Session hijacking with CloudWatch
- 17:54Data loss prevention with Macie
- 21:26Automating fraud detection
- 24:38Real-time log analysis architecture
- 32:26Fraud detection with Spark and EMR
- 38:08Conclusion and Q&A
related talks
28mPatterns for scalable security detections on AWS
18mHow AWS Partners use observability to strengthen customer security
20mFantastic AWS Attacks and Where to Find Them
33mDetecting SSRF events on AWS using Splunk
39mHow Security Incidents Are Different ... and How They're Exactly the Same
47mBreaking the Cloud to Rebuild It: A Tale of 3 ☁️ Breaches!
55mStreamline security analysis with Amazon Detective
50mTo Normalized Logs, and Beyond - Building a Threat Detection...