Ravelin
Ravelin Raises $44M Series A for Offensive Agent Red-Teaming That Continuously Stress-Tests Customer-Facing AI Systems
Red-teaming agents that probe customer-facing AI for prompt injection, tool abuse, and data exfiltration paths—continuous campaigns with fix tickets, not annual pen-test PDFs.
Soo-yeon Park · ex-CrowdStrike
“You cannot govern agents you have never attacked. We industrialize the adversary for every release train.”
“Agent security is dual-use by nature—Ravelin productizes the defensive side of that arms race.”