Remediation-Aware Reachability: Patching Containers, Prioritizing with Agentic-CTI, and Scaling...
summary
Platform-grade pipeline that fuses real-time evidence with prediction and reachability so teams act on what is burning now, not CVSS noise: an open feed called Phoenix Blue (ingests 15+ sources), a neural classifier and judge chain that processes millions of advisories hourly, an OSS package risk score and a 77‑parameter model to flag malicious packages, agentic LLM enrichment with adversarial validation to catch hallucinations, and automated library remediation gated by reachability that can push fixes in 15 minutes.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 3:142024 threat landscape
- 7:29Why traditional VM fails
- 13:36Vulnerability prioritization questions
- 17:29Vulnerability to remediation
- 21:00AI for triage and prevention
- 24:54Phoenix Blue threat intel
- 37:17Real-time prioritization scoring
- 43:17Fighting AI with AI
- 46:44Conclusion and Q&A
related talks
40mApplying Artificial Intelligence to the Incident Response Function
30mIntelligent Pipelines: Bringing AI/ML into DevSecOps Automation
51mPractical Approach to Automate the Discovery & Eradication of Open-Source Software Vulnerabilities
59mMadness of Vulnerability Management in Modern Cloud, Container, How to Win the Battle...
29mContainer vuln management with (hopefully) minimal...
30mTopics of Interest: Automatic Vulnerability Remediation: The Trusted and Secure Road ...
13mBeyond Manual Firefighting
36mUsing AI to Discover Silently Patched Vulnerabilities in Open...