General
August 25, 2024

LLM case studies to get Security Teams prepared

By venkat

Enterprises are going to adopt a wide range of Generative AI use cases. The security teams need to understand them, and how they are built so they can eventually protect it better. Here we are learning from the perspective of Incident Response to cyber attacks on the AI platforms.

Case study 1 – A great CTI tool – Neutocti

This project gives a great insight into how we can understand the typical patterns of Generative AI that can be used for security use cases. It can be adapted to extend your specific use cases.

Concepts: RAG, small domain-specific custom models, Fine tuning

Case study 2 – A sample Councilor Co-Pilot project that was built in 24 hours

This project gives you an example to build something for your organisation so you can get familiar with the basics of AI.

Concepts: RAG, Question Answer system with context

  • https://devpost.com/software/counselor-copilot
  • https://github.com/zrizvi93/trevorhack

About The Author

Venkat is founder of FlexibleIR. He brings 20 years of experience in building tools and products at Sun Microsystems, Intel, Novell, HP, Yahoo,Tesco and startups. He has developed test suites and frameworks for post silicon validation of the Xeon processor family (Fuzzing). He has worked deeply on UFS files system at SUN Microsystem. Was a security paranoid at Yahoo.