Empirical Security

About Empirical Security

Empirical Security is a cybersecurity intelligence company that develops artificial intelligence models tailored to an enterprise's specific infrastructure and threat landscape. Founded in 2024 by Michael Roytman and Jay Jacobs, the Chicago-based firm emerged from stealth in July 2025 with $12 million in seed funding. The founding team was joined by Ed Bellis, who assumed the role of CEO, reuniting the leadership trio from Kenna Security, a risk-based vulnerability management pioneer acquired by Cisco in 2021. Roytman, now CTO, and Jacobs, Chief Data Scientist, were instrumental in creating the Exploit Prediction Scoring System (EPSS), an open-source standard for predicting vulnerability exploitation.

The company's core offering is a dual-model AI architecture designed to provide more precise threat prioritization than generic, one-size-fits-all security tools. This system combines global models, which are trained on millions of daily exploitation events from internet-wide datasets, with local models. These local models are trained on an organization's unique data, systems, and technical stack, creating a customized defense that adapts to the specific environment. This approach enables security teams and CISOs to make more accurate, evidence-based decisions on which vulnerabilities to address first, reducing alert noise and focusing resources on the most critical threats.

The business model targets enterprise security teams, offering a platform to enhance their vulnerability management programs. The recent seed round, led by Costanoa Ventures with participation from DNX Ventures, Sixty Degree Capital, and HPA, will be used to accelerate product development and go-to-market efforts. The involvement of notable angel investors from companies like Google/Mandiant and Qualys underscores confidence in the company's approach to personalized AI in cybersecurity.