Semgrep eventsAdding AI to security analysis doesn’t automatically improve outcomes. In many environments when done haphazardly, it simply increases cost, complexity, and risk.
These lightning talk sessions explore the architectural and operational decisions involved in combining LLM reasoning with traditional static analysis systems at scale. Topics include benchmark design, orchestration workflows, validation strategies, false-positive reduction, and the realities of building systems that meet actual production demands.
Tools + Reasoning = Agents
What the data shows for how teams use AI Memories to inform the context window
Methodology for Benchmarking LLM + SAST performance.