Kevlar, a French Cybersecurity Startup, Tackles SOC Challenges with AI-Driven Solutions

🎬 The video features a discussion with Ahed, co-founder of Kevlar, a French cybersecurity startup addressing challenges faced by Security Operations Centers (SOCs), particularly alert fatigue. SOCs struggle with overwhelming alert volumes, low automation levels, and static playbook-based tools, leading to burnout and reduced defensive posture. Kevlar’s solution uses AI-driven autonomous investigation, combining graph-based orchestration with specialized machine learning models (including LLMs and computer vision) to mimic human SOC analysts, reducing investigation time to approximately 2.5 minutes per alert. The platform is source-agnostic, integrates with existing EDRs, SIEMs, and threat intelligence tools, and operates as a SaaS model with SOC 2 compliance. Kevlar has been in development for three years, employs 47 people, and targets both enterprises and MSSPs, claiming measurable improvements in alert triage and analyst efficiency. The company emphasizes European sovereignty but serves clients in France, Germany, Sweden, the UK, and the US, with plans to expand further.