Zak Krider, Trellix's Director of Strategy and AI, shares how Trellix has successfully integrated generative AI into their security operations and democratized access to AI models across the organization.

Topics Include:

• Trellix formed from McAfee Enterprise and FireEye merger
• Provides full security stack visibility in single platform
• Serves SMBs to Fortune 500 and government customers
• Used machine learning for two decades with 30 models
• Recently pivoted to generative AI with Wwise platform
• AI finds critical events among thousands daily alerts
• Incorporates threat hunting knowledge into AI prompt structures
• AWS Bedrock central to AI strategy for model flexibility
• Formed small tiger team to investigate generative AI
• Anthropic Claude provided breakthrough "aha moments" for capabilities
• Adopted "fail fast, learn fast" innovation culture approach
• Enabled employee access to models through Bedrock API
• Conducted innovation jam sessions with VC-style pitches
• AI decoded Base64 without prompting, identified benign activity
• Junior analysts elevated to level two with AI
• Common misconception: models train on customer data falsely
• Early challenge: providing too much data overwhelmed models
• Smaller models hallucinated more with plausible-sounding responses
• Design partner programs help prioritize product development
• Democratize AI access beyond just technical teams
• Test multiple models for specific use cases
• Large models work better than small ones initially
• Prompt engineering crucial for effective model communication
• Model Context Protocol will gain traction next year
• Backend data security remains largely unsolved challenge
• Federal customers require on-premises, air-gapped AI solutions

Participants:

• Zak Krider – Director of AI and Innovation, Trellix

Further Links:

• Website: https://www.trellix.com
• Trellix on AWS Marketplace

See how Amazon Web Services gives you the freedom to migrate, innovate, and scale your software company at https://aws.amazon.com/isv/