Government cyber
Former NSA Technical Director
Served in technical leadership roles supporting computer network operations and defensive security work.
KevinBytes
Cybersecurity research and writing
Kevin O'Connor
Cybersecurity researcher focused on AI security, threat research, and high-assurance systems.
Kevin O'Connor is a former NSA Technical Director who worked on CrowdStrike's Advanced Research team and led threat research at Adlumin, now part of N-able. Current work centers on adversarial AI systems, detection engineering, and practical security tooling.
Third-party proof
Source-linked public record
SecurityWeek/Media profile
Hacker Conversations: Kevin O'Connor
A third-party profile covering Kevin O'Connor's path through offensive and defensive cybersecurity work.
Source (new tab)
Security Magazine/Authored analysis
Holiday weekend cyber threat preparation
An authored article on holiday-weekend cyber threat preparation for security teams.
Source (new tab)
CISO Tradecraft/Podcast recap
SMB CISO challenges
Episode recap featuring Kevin O'Connor, then Director of Threat Research at Adlumin, on SMB security leadership and MDR constraints.
Source (new tab)
Cybersecurity Summit/Speaking reference
Public security summit listing
A stable third-party event surface that helps connect the public profile to security-industry work.
Source (new tab)
Background
Technical security work across government, research, and operations
Threat intelligence
CrowdStrike Advanced Research
Worked on advanced research and threat-intelligence problems at commercial scale.
Detection engineering
Adlumin / N-able threat research leadership
Led threat research focused on ransomware, intrusion activity, and practical detection outcomes.
Current work
Founder of TKOResearch
Builds and publishes security research, practical tools, and advisory work through TKOResearch and KevinBytes.
AI security work
Agentic systems, adversarial testing, and security boundaries
AgentKillChain
Public work on evaluating autonomous AI agents against adversarial patterns such as prompt injection, memory poisoning, tool misuse, and context drift.
High-assurance security engineering
Research and advisory focus on systems where AI behavior, tooling, and security boundaries need to be examined before deployment.
Writing and tools
Writing and practical tools
Security tools
A compact index of practical browser-based utilities for security analysis and technical workflows.
GHS label builder
A canvas-based chemical label tool for structured hazard communication outputs.
NFPA diamond builder
A practical NFPA 704 placard builder for quick safety-label generation.
Contact