Qevlar AI: Why We Invested

Autonomous AI to Redefine Security Operations

I’m thrilled to announce that Forgepoint Capital International is co-leading Qevlar AI’s new funding round alongside EQT Ventures, bringing the fast-growing startup’s total funding to $14M.

From my earliest conversations with co-founders Ahmed Achchak and Hamza Sayah, it was clear that Qevlar AI‘s vision to redefine the Security Operations Center (SOC) with AI perfectly aligns with Forgepoint’s investment thesis at the convergence of AI and cybersecurity.  

Manual Triage, Alert Fatigue, and Imperfect Automations
The current state of the SOC is unsustainable. SOC teams rely on an ever-expanding set of tools to respond to the relentless influx of endpoint, network, identity, data, and cloud threats. Analysts are overwhelmed by a massive volume of alerts that require manual triage but lack actionable context. The situation is exacerbated by the ongoing cybersecurity talent shortage. 

SOC automation is in high demand, but existing solutions come with tradeoffs. Legacy Security Orchestration Automation and Response (SOAR) and low-code automation solutions use rule-based playbooks that require time-consuming configurations and are often insufficient in the face of the unpredictable threat landscape. AI copilots and LLM-based SOC automation platforms are more flexible and dynamic but lack explainability, reliability, and consistent outputs.    

Autonomous and Explainable AI-Led SOC Investigations
Enter Qevlar AI, which enables completely autonomous and explainable SOC investigations without playbooks, configurations, or prompts. Autonomous AI agents enrich and analyze alerts from existing SIEM, SOAR, EDR, CTI, and ticketing systems, generating comprehensive incident reports and recommending remedial actions- all in around three minutes. The SOC becomes more productive and proactive as analysts gain time to perform higher level incident response, threat hunting, and posture management activities. Alert classification outcomes improve as well: a recent benchmark study of real-world email alerts found that Qevlar AI achieved 99.8% accuracy with just 0.3% false positives and 0.1% false negatives at 5 minutes per investigation, an improvement in every metric compared to expert SOC analysts. 

Qevlar AI’s ‘smart AI’ technology is highly differentiated and delivers unmatched intelligence, orchestration, and reasoning to reduce false negatives and false positives. Proprietary knowledge graph models enable sophisticated alert analysis and identify hidden relationships between entities to pinpoint kill chain steps unique to each customer’s environment, while LLMs manage features, reporting, and outputs. Analyst oversight is only required for outcome validation and remediation. This proprietary architecture also strengthens customer data security.   

The results speak for themselves. Qevlar AI’s customers including Nomios, Global Connect, Equans, and Almond experience 99.8% accuracy, 90% conclusive investigations, and up to 10x increased SOC efficiency.

An Innovative Team Backed by the Best of the Best
Just as impressive as Qevlar AI’s technology is its team. When I first met Ahmed and Hamza during VivaTech 2024 in Paris- where Qevlar AI would go on to win Orange Cyberdefense’s cybersecurity startup competition– I was deeply impressed by their extensive AI expertise, strong communication and coachability, drive, and global ambitions.  

Ahmed and Hamza have rich backgrounds in AI, machine learning, and automation. Hamza was previously a data scientist at Plüm Énergie (acq. Octopus Energy) and the first employee and Head of Research at AI-automated code testing startup Ponicode (acq. CircleCI). Ahmed consulted with enterprise banks at BCG before developing machine learning algorithms at global financial services firm Natixis and becoming CTO at auto parts reseller Minautor.  

In response to Qevlar AI’s explosive growth after less than two years in the SOC automation market, the team has quickly grown from 4 to 25+ with recent additions including seasoned product, sales, and marketing leaders to support go-to-market objectives. The expanding team’s combination of technical and commercial expertise is a winning combination.  

Qevlar AI is backed by a deep bench of successful entrepreneurs, investors, and enterprise partners. Supporters include previous Seed round leader EQT Ventures and angel investors Olivier Pomel (CEO of Datadog), Florian Douetteau (CEO of Dataiku), and Mehdi Ghissassi (ex-Director of Product at Google DeepMind) in addition to an advisory board with strong cybersecurity operations expertise. The company has also been selected to several high-profile startup accelerators including Meta and Hugging Face’s joint AI startup program, Microsoft’s GenAI Studio, and Station F’s Future 40. Forgepoint Capital International is honored to be a part of this exceptional community of support as we work together to help propel Qevlar AI forward at scale.  

Looking to the Future
Qevlar AI has a rare and valuable set of advantages- the people, technology, market opportunity, and support to become a global leader in AI-enabled SOC automation. We look forward to working alongside Ahmed, Hamza, and the team as they use AI to automate and transform the SOC.