Qevlar AI Autonomous SOC Platform

Qevlar AI is a Paris-based company founded in 2023 by Ahmed Achchak (CEO) and Hamza Sayah (CTO), AI engineers who saw security operations centers drowning in alert noise. It builds an autonomous AI SOC platform for enterprises and managed security service providers.

When an alert fires in a SIEM or EDR, the platform enriches it with internal and external data, correlates related alerts into one incident story, decides whether it is malicious, writes a report with indicators of compromise and step-by-step reasoning, and recommends containment or tuning. A graph orchestrator controls decisions, with LLMs used only for narrow tasks such as enrichment and summarization, so every verdict is transparent. It also runs autonomous threat hunts and links confirmed exploitation to vulnerability remediation. It is offered as SaaS or private cloud and is used by Orange Cyberdefense, Atos, Sodexo and Mercedes-Benz.

• Autonomous investigation: Enriches, correlates and investigates every SIEM and EDR alert without playbooks.
• Incident story: Groups related alerts into one incident and maps blast radius.
• Explainable reports: Lists indicators of compromise, reasoning steps and remediation actions.
• Threat hunting: Hunts for emerging threats, TTPs and active exploitation.
• SOC and vulnerability alignment: Raises assets with confirmed exploitation in the remediation queue.
• Flexible deployment: SaaS or private cloud, headless via ticketing or as a dedicated console.

A Vietnamese MSSP can plug the platform into its SOC to investigate client alerts from Microsoft Sentinel or Splunk around the clock. A bank SOC can auto-close false-positive phishing and identity alerts and focus analysts on confirmed incidents. An enterprise can link exploited assets to its vulnerability patching queue.

Average alert investigation time drops to about 3 minutes, which the company describes as a 10x reduction. Up to 80% of tickets are closed automatically while 100% of alerts are investigated around the clock. Transparent step-by-step reports let analysts validate verdicts quickly. Investigation knowledge stays in the platform, reducing dependence on individual analysts.

MSSPs face alert overload as client numbers grow, and AI agents that investigate tier-1 alerts are becoming a way to expand capacity without hiring. European MSSPs such as Orange Cyberdefense and Nomios have already integrated such platforms. Southeast Asian and Vietnamese MSSPs serving banks and SMEs face similar pressure from analyst shortages.

Founded in 2023 in Paris, France. Raised 14 million USD led by EQT Ventures and Forgepoint Capital International, then 30 million USD in March 2026 co-led by Partech and Forgepoint. Holds SOC 2 Type 2. Awards include MSP Today Product of the Year 2025 and InCyber Forum Europe Growth Award 2026. Nomios is its first European integrator and reseller.