At Hordanso LLC, I test LLMs, AI applications, and autonomous agents for prompt injection, jailbreaks, data leakage, and unsafe tool use. I use Garak, PyRIT, and TextAttack to design adversarial tests and assess agent communication paths, including MCP and A2A.
I map AI security findings to NIST AI RMF and ISO/IEC 42001, and build automated stress tests that probe agents for unauthorized tool calls and unsafe actions. I document findings with severity ratings and remediation guidance, working with security and engineering teams to close vulnerabilities.
Previously, I participated in red team engagements across hybrid Active Directory and cloud environments at Hordanso LLC, and conducted offensive security work at Texas Digital Pros. In earlier cloud roles, I co-designed AWS security controls and contributed to a 25% reduction in cloud infrastructure costs at SkySpace Global.

