Empirical Security Raises $12M to Stop Attacks with Custom Cybersecurity AI Models

Ex-Kenna Security founders and Exploit Prediction Scoring System (EPSS) pioneers the future of data-centric, customized cybersecurity defenses

CHICAGO, July 17, 2025 /PRNewswire/ -- Empirical Security, the cybersecurity company enabling organizations with Custom AI models, today announced $12 million in seed funding. The round was led by Costanoa Ventures with participation from DNX Ventures, Sixty Degree Capital, HPA, and strategic investors Jonathan Cran, founder of Intrigue (acquired by Google/Mandiant), Wade Baker, creator of the Verizon Data Breach Investigations Report and founder of The Cyentia Institute, and Gerhard Eschelbeck, former CTO of Qualys and CISO of Google.

Kenna Security co-founder Ed Bellis also joins Empirical Security as the Chief Executive Officer (CEO), reuniting with Empirical Security's co-founder and Chief Technology Officer (CTO), Michael Roytman and co-founder and Chief Data Scientist, Jay Jacobs.

Companies regularly use generic, global models to prioritize which of the hundreds of daily threats to pay attention to. But every company has its own unique combination of data, systems and technical stack. Empirical uses that data to build and train custom 'local models' that make more accurate predictions about the most critical threats.

"Today's cyber attacks are custom-built using AI and your own infrastructure against you," said Michael Roytman, Co-founder and CTO of Empirical Security. "Defending with generic, one-size-fits-all models is a start, but only custom, localized models - trained on your data and environment - can close that gap."

Empirical Security's dual model architecture allows security teams to leverage the best of both global and local models:

• Global models trained on nearly 2 million daily exploitation events mined from internet-scale datasets to detect and anticipate the latest attacks.
• Local models trained to adapt the intelligence to each customer's unique infrastructure, operations, and threat profile.

This approach delivers precise, measurable predictions and decision support, empowering CISOs to act confidently and justify cybersecurity decisions with evidence-based risk analysis.

"We backed Ed, Michael, and Jay at Kenna, where they pioneered the risk-based vulnerability management movement," said John Cowgill, General Partner at Costanoa Ventures. "With Empirical, they're doing it again—but replacing generic risk scores with local AI models that tailor scores to each enterprise. We're thrilled to partner with them on what we believe is a 10x bigger opportunity to transform how security teams prioritize and act on risk."

To learn more about Empirical Security, please visit here.

About Empirical Security
Empirical Security is a cybersecurity intelligence company specializing in building AI and machine learning models that empower data-driven security decisions. Their state-of-the-art models are designed for rapid fine-tuning in each customer deployment, leveraging curated, AI-mined, and customer-provided data. This approach delivers localized models offering advanced decision support tailored to each organization's environment, risk threshold, and specific needs.

About Costanoa Ventures
Founded in 2012, Costanoa Ventures partners with builders as early as company formation, with a focus on apps and infrastructure in data, dev, security and fintech. Costanoa is a long-term, boutique partner to entrepreneurs from the earliest stages of company building with expertise from its BuilderOps team. For more information, please visit www.costanoa.vc.

SOURCE Empirical Security, Inc.