Grant Abejar
Cybersecurity/AI Practitioner
Securing the digital frontier, one vulnerability at a time.
Right now
By the numbers
Practice areas
Capabilities
Avg proficiency
01 — About
The short version
Passionate cybersecurity professional with expertise in penetration testing, security architecture, and incident response. I combine technical skills with a deep understanding of threat landscapes to protect organizations from evolving cyber threats.
When I'm not hunting vulnerabilities, you'll find me contributing to open-source security tools and sharing knowledge with the community.
Offense-informed defense
Pentesting and threat modeling that feed directly into hardening and detection.
AI with guardrails
LLM integrations built with prompt-injection testing and data isolation from day one.
Automation first
If it happens twice, it gets scripted — Python, Bash, and CI pipelines.
02 — Skills
What I work with
Security
- Penetration Testing
- Vulnerability Assessment
- SIEM & Log Analysis
- Incident Response
- Threat Modeling
- Network Security
Development
- Python
- Bash/Shell
- JavaScript/TypeScript
- Go
- SQL
AI & Machine Learning
- LLM Integration
- Prompt Engineering
- AI Security
- ML Ops
- Neural Networks
Tools & Platforms
- Burp Suite
- Metasploit
- Wireshark
- Nmap
- Docker
- AWS Security
03 — Credentials
20 certifications.
04 — Work
Selected projects.
contextual-secrets-scanner
Contextual secret scanner: regex + entropy + LLM-grounded context filtering to cut false positives.
jailbreak-bench
LLM jailbreak/prompt-injection benchmark: curated test corpus, deterministic scoring, LLM-as-judge evaluator.
phish-payload-gen
Defender-side phishing-awareness training toolkit: 9-rule detector, difficulty grader, defanger, LLM analyst explainer, auth-gated lure generator.
insider-threat-detector
Behavior-analytics insider threat detector: per-user profile, 8 signals, weighted aggregation, LLM-grounded explainer.
vpc-baseline-engine
Read-only AWS VPC posture/baseline checker (boto3 + JSON loaders, 10 bundled rules, LLM explainer)
ddos-sig-classifier
Hybrid (signatures + ML) DDoS classifier on NetFlow-style records, with LLM-grounded explanations.
ai-waf-anomaly
HTTP WAF anomaly-detection layer: regex signatures + z-score/iForest/path-rarity, hybrid allow/log/challenge/block decision, LLM explainer
cve-triage-engine
Risk-aware CVE triage: CVSS v3.1 + EPSS + CISA KEV + asset registry, with LLM explainer (citation-guarded)
adv-ml-toolkit
Adversarial ML toolkit: vision+text attacks, input-space defenses, robustness benchmark, LLM explainer with citation guard
recon-agent
LLM-driven defensive recon orchestrator (nmap/http/ffuf) with hard scope and tool whitelists