Not another chatbot: How Trellix virtual analyst automatically scales
summary
Trellix Wise uses Amazon Bedrock plus data lakes, S3 and indexing to auto-investigate alerts with 20 Q&A per alert in sub-second, turning noisy detection into decisioning that flags only the few humans must see. Average customer ingests ~1 billion events and ~1,000 alerts/day (≈65 events/alert) and Bedrock-driven triage narrows that to about 10 actionable alerts per analyst per day. Prebuilt question/answer investigations spot demo accounts, Tor activity and cloud hosting IPs, allow natural‑language tuning, and runs in production today.
AI-written summary · about
chapters (11)
- 0:00Introduction to AI in Security
- 1:21The Shift to Generative AI
- 1:41How Auto-Investigation Works
- 1:51Components of the AI Solution
- 2:31Real-world Scenario Simulation
- 3:10Efficiency Gains with AI
- 4:19Example: Investigating a User
- 4:58Real-World Threat Analysis
- 8:06Scaling the Security Team
- 8:54Tuning the AI Ecosystem
- 9:27The Future of Threat Detection
related talks
30mGenerative AI in action: from POC to business value
12mSecurity data gravity: Visibility & threat detection for data lakes
43mBeyond Chatbots: How Amazon Bedrock Agents Drive Enterprise AI
01h 00mData is out, knowing is in: Applying AI to automate cloud operations
18m1,200 analytics requests a day: Clay and Vercel's Bet on AI-Native Analytics
13mExploring AI-Powered Chatbots and Their Transformative Role
49mFight Smarter: Accelerate Your SOC with AI Discovered Attacks
34mAutomating SRE Operations with Multi-Agent AI