Research Seminar: Edwin Kairu

April 27, 2026

12:30 p.m. CAT

A203

Building Autonomous AI for Security Operations 

Can a language model run a penetration test? Investigate a SIEM alert? Preserve forensic evidence without a human in the loop? This talk presents a working deployment of three collaborating AI agents–an orchestrator, a security node, and a Kali-based pentester–built at CMU-Africa on open-source infrastructure. Edwin Kairu walks through the five design principles that separate useful agents from demoware: identity, skills, memory, safety guardrails, and access control. A live demonstration launches the pentest agent mid-talk and lets it work autonomously in the background while the presentation continues.

Bio

Experienced cyber security professional and entrepreneur with a demonstrated history of improving the security posture of multi-ple organizations. Skilled in security reviews, intrusion detection, incident response, vulnerability management, data protection and network forensics. Ardent professional with over 15 years of industry experience, a Masters in Information Security Policy & Management from Carnegie Mellon University, and a Bachelors in Business Administration (Management Information Systems) from the State University of New York.

Upcoming Events