Talk at the oop Conference for Software Architecture
Do traditional tools like SAST or SCA have a future thanks to AI?
About the talk
Although modern development tools reduce the risk of unintentional security vulnerabilities, security remains a highly relevant topic in companies. The use of powerful AI systems offers new ways to optimize and automate security processes.
In this session, we will show how traditional scanning tools (SAST, SCA) can be combined with the capabilities of Large Language Models (LLM) to:
- Reduce false positives and negative positives,
- Focus on critical security findings,
- Improve the analysis and assessment of findings,
- Effectively deploy a greater number of specialized tools.
We compare thousands of findings analyzed manually and by LLM and highlight the differences between free and proprietary models.
Target audience: Developers, architects, security managers
Prerequisites: A basic understanding of common code vulnerabilities is required.
Difficulty level: Advanced
Time and Location: Munich, Thursday, February 6, 2025 – 09:00 – 10:30
