Security operations startup Brava Security has launched publicly after raising $15 million in seed financing. The round was led by F2 Venture Capital and Pitango VC, with additional participation from angel investors.
Founded in 2024, the Tel Aviv-based company is led by chief executive Idan Basre and chief technology officer Itay Gabbay, both described as alumni of Israeli intelligence organizations. Brava maintains research and development operations in Israel and a commercial presence in the United States.
Attack simulation and telemetry analysis
Brava’s platform uses AI agents to model adversary activity in a customer’s specific environment. Rather than only testing whether a detection rule exists, the company says its technology simulates the telemetry an attack would create across an organization’s log sources. The objective is to identify gaps where security tools may fail to observe or alert on malicious behavior.
According to Brava, the platform incorporates business context to show security teams how a particular attack path could appear in their infrastructure. It can also help analysts investigate findings and suggest changes to SIEM detection logic.
The company argues that validating the availability and quality of evidence is an important prerequisite for more automated security operations. AI-based investigation tools cannot reliably assess incidents that are not represented in the logs and alerts available to them, Brava said.
Integrations and expansion plans
Brava says customers can begin receiving initial findings within 10 to 15 minutes of onboarding. The company said its product connects with more than 70 data sources and destinations and is designed to work without requiring organizations to replace their existing security infrastructure.
The new capital will be used to further develop the product and expand its market presence. Brava enters a growing field of cybersecurity vendors applying generative and agentic AI to exposure validation, attack simulation, detection engineering, and security operations workflows.
For enterprises, the company’s approach is intended to provide a more continuous way to test whether defensive tooling captures the signals needed to detect real-world attacker techniques before an incident occurs.
