Skip to main content

Empirical Security

Category: AI in Cybersecurity

AI-driven cybersecurity exposure management platform that builds custom predictive models for enterprise vulnerability prioritization Empirical Security was founded in 2024. The company is led by Ed Bellis. Based in Chicago, Illinois, USA. Team size: 11-50. Total funding raised: $37M. Latest round: Series A. Key investors include Brightmind Partners, Costanoa Ventures, Hyde Park Angels (HPA), DNX Ventures, Sixty Degree Capital.

Founded
2024
Headquarters
Chicago, Illinois, USA
Team size
11-50
Total funding
$37M

Value proposition

Empirical Security builds precision AI models — both global (Foundation) and organization-specific (Radiant) — that predict which vulnerabilities will be exploited in a given enterprise environment, enabling security teams to prioritize remediation based on real-world exploitation data rather than static severity scores.

Products and solutions

Foundation Model (global predictive model monitoring 18,000+ exploited CVEs with near-real-time exploitation telemetry), Radiant Model (custom local model trained on each organization's assets, configurations, and internal telemetry), EPSS Model (Exploit Prediction Scoring System — open-source ML model predicting exploitation probability), AI Agent for workflow automation, API and MCP access

Unique value

Only exposure management vendor that publishes false positive rates and measures performance; dual-model architecture combining global exploitation telemetry with organization-specific local models; co-creators of EPSS (industry-standard exploitation prediction scoring system used by Tenable, Qualys, CrowdStrike, Microsoft, Wiz)

Target customer

Enterprise security teams and CISOs at large organizations, particularly those overwhelmed by alert fatigue from generic vulnerability scoring systems

Industries served

Cybersecurity / Vulnerability Management / Exposure Management

Technology advantage

Dual-model architecture (global Foundation model + local Radiant model); near-real-time hourly exploitation telemetry across 18,000+ CVEs (12x CISA KEV catalog); team invented Risk-Based Vulnerability Management category at Kenna Security and co-created EPSS; publishes transparent performance metrics (false positive rate, coverage, efficiency)

How they differentiate

Unlike competitors that use generic one-size-fits-all scoring (CVSS) or global-only models, Empirical trains organization-specific local models (Radiant) that incorporate each enterprise's unique assets, configurations, controls, and telemetry. The company also publishes transparent performance metrics including false positive rates — unique in the exposure management space.

Main competitors

Tenable, Qualys, CrowdStrike (Falcon Exposure Management), Wiz

Key partnerships

EPSS maintained in partnership with FIRST.org, EPSS embedded by Tenable, Qualys, CrowdStrike, Microsoft, and Wiz, integrations with Qualys, Tenable, Rapid7, AWS, Azure, GCP, ServiceNow

Notable customers

Leading enterprises (specific names not publicly disclosed), EPSS used by Tenable, Qualys, CrowdStrike, Microsoft, Wiz

Major milestones

2024: Company founded by Ed Bellis, Michael Roytman, and Jay Jacobs, 2025-07: Emerged from stealth with $12M seed round led by Costanoa Ventures, 2026-07: Raised $25M Series A led by Brightmind Partners ($37M total), Maintains and powers EPSS (Exploit Prediction Scoring System) for FIRST.org

Market positioning

Pioneer of next-generation AI-native vulnerability management, building on the founders' legacy of creating the Risk-Based Vulnerability Management category at Kenna Security. Positioned as the intelligence layer that makes existing security tools smarter through custom predictive models.

Geographic focus

United States (Chicago-based), serving global enterprise customers

About Ed Bellis

Ex-Co-Founder & CTO of Kenna Security (acquired by Cisco in 2021); Former CISO of Orbitz; Security roles at Bank of America, CSC; Co-authored "Modern Vulnerability Management" and contributed to "Beautiful Security"

Latest news about Empirical Security

More AI in Cybersecurity companies

Official website: