Brava Security, a Tel Aviv-based cybersecurity startup, has emerged from stealth with $15 million in seed funding. The company is developing AI technology that tests whether an organisation's security systems would actually detect real attacks.
The round, announced on Wednesday, November 4, was led by F2 Venture Capital and Pitango. Angel investors from Anthropic, Microsoft and Check Point also took part.
Brava was founded in 2024 and has about 20 employees. Despite its early stage, the company says it already counts Fortune 100 companies among its customers.
The problem: untested detections
Large organisations rely on security information and event management (SIEM) systems and other tools to detect cyberattacks. These systems collect data, known as telemetry, from across an organisation's networks, devices and applications, and apply detection rules to identify suspicious activity.
The challenge is that many of these detection rules are never fully tested. Security teams write rules designed to catch specific types of attack, but they often cannot confirm whether those rules will trigger when a real attack occurs. Changes to systems, data sources or configurations can also break detections without anyone noticing.
As a result, organisations may believe they are protected against certain threats when in fact their systems would miss them. That gap can leave attackers free to operate undetected.
How Brava works
Brava's platform uses multiple AI agents to address this problem. Instead of running an actual attack, which could be risky and disruptive, the system generates the telemetry that a real attack would produce. That simulated data is then fed into the organisation's security tools to check whether their detections fire as expected.
This approach allows organisations to validate their detections safely and continuously, without exposing their systems to real attacks. The company says its platform can produce insights within 10 to 15 minutes.
Brava integrates with more than 70 security tools, including widely used platforms such as Splunk, Microsoft Sentinel, Google Security Operations and CrowdStrike. That breadth of integration allows it to work with the security tools that large enterprises already use.
Proving evidence before automation
Co-founder and chief executive Idan Basre has framed Brava's mission around the need for proven evidence in security. "You can't automate a decision based on evidence no one ever proved," he said.
The point is particularly relevant as security teams adopt AI to automate more of their work. AI systems are increasingly used to triage alerts, investigate incidents and even respond to threats automatically. But those automated decisions depend on the quality of the underlying detections. If the detections are flawed, automation can amplify the problem.
Brava's approach aims to give organisations confidence that their detections work before they rely on them for automated responses.
The founders
Brava's founders bring deep experience in cybersecurity and AI. Basre previously served as a product lead for Microsoft Defender, one of the world's most widely used security products. He also spent 10 years in Unit 8200, the Israeli military's intelligence unit, and is a graduate of Talpiot, an elite Israeli military programme that trains technology leaders.
Co-founder Itay Gabbay previously led joint AI research with DARPA, the US Defense Advanced Research Projects Agency. His experience in advanced AI research underpins the company's multi-agent approach.
