Skip to content

Evoke Security

Pre-seed platform for continuously inventorying, threat-modeling, and monitoring AI agents across the enterprise.

Visit Website ↗ + Add to Compare
52/100Incremental Innovator

Overview

Evoke Security is building a security platform focused specifically on the “agentic workforce” — the AI agents enterprises are increasingly giving access to sensitive data and systems. Its platform continuously inventories which AI agents exist across an environment, threat-models what they can access and do, and monitors their behavior, aiming to close a visibility gap that most existing security tooling wasn’t built to cover.

The Austin, Texas company was founded in 2025 by CEO Jason Rebholz, a former CISO and incident-response leader, and CTO Jeff Chan. Evoke raised a $4 million pre-seed round in February 2026 led by Crosspoint Capital Partners with participation from Red Cell Partners, and was selected for the 2026 CrowdStrike, AWS, and NVIDIA Cybersecurity Startup Accelerator as well as the NVIDIA Inception Program — a competitive, independently juried selection process among a large accelerator applicant pool.

The focus on agent inventory and threat modeling, rather than general-purpose prompt filtering, is a sensible niche given how fast agentic AI deployment is outpacing governance. But Evoke is genuinely early: pre-seed funding, a 2025 founding date, and no public customers or case studies yet. The accelerator selections are a meaningful vote of confidence from credible technical partners, but they are not evidence of production efficacy.

Innovation Matrix Assessment

Innovation Velocity 6/10

Raised pre-seed funding and secured two competitive accelerator placements (CrowdStrike/AWS/NVIDIA, NVIDIA Inception) within roughly a year of founding.

Operational Value 6/10

Targets a real, emerging blind spot — enterprise visibility into what AI agents can access — but is unproven in production at this stage.

Market Momentum 4/10

Only $4M in pre-seed funding; accelerator selection is a promising signal but not evidence of customer adoption.

Category Disruption 5/10

A timely focus area, but numerous competitors are racing into agentic AI security posture simultaneously, so it's too early to call this category-defining.

Real-World Efficacy 3/10

No customers, case studies, or independent testing were found; the company is pre-launch relative to public evidence.

Enduring Relevance 7/10

Agentic AI inventory and governance will only become more necessary as enterprises deploy more autonomous agents.

Why CISOs Should Care

Would give security teams a running inventory of what AI agents exist and what they can touch — visibility most organizations currently lack entirely.

What Makes It Different

Focused specifically on agent inventory and threat modeling rather than general LLM prompt/output filtering.

The Matrix Verdict

52/100 — INCREMENTAL INNOVATOR

A credibly backed, very early-stage bet on agentic AI visibility; Emerging/Unranked-adjacent territory reflected in a conservative Incremental score pending real production evidence.

Editorial Note: Claims vs. Verified Findings

Funding and accelerator selections are independently reported. All product-effectiveness claims are vendor-stated; no customers or case studies were publicly available at the time of research.

Sources