Angelo Schranko de Oliveira

About Angelo Schranko de Oliveira

I build AI systems. I break them. AI/ML for Security:Designing, developing, and deploying production-grade Machine Learning (ML), Deep Learning (DL), Graph Neural Networks (GNN), and Large Language Models (LLM) systems for SaaS security, turning large-scale telemetry into detection, correlation, UEBA, and real-time threat response. Securing AI/ML (AI Red Teaming):Actively attacking AI/ML systems and LLM applications: prompt injection, adversarial ML (evasion, data poisoning), model abuse, and pipeline exploitation. If it can be broken, I’ll find how. AI/ML-Assisted Low-Level Windows Vulnerability Research:Applying AI/ML to Windows user-mode and kernel exploitation: fuzzing, heap corruption, ROP chain synthesis, exploitability analysis/scoring, and large-scale crash triage. Lift heavy. Play Doom Metal bass. Repeat.

Experience

1. Subject Matter Expert

   CompTIA

   May 2018 — Present · Chicago, IL, US

   cybersecurity certifications, developing exam objectives and assessment content aligned with real-world threats and practices.I’ve participated in SME workshops for PenTest+, CySA+, and Cyber AI+, helping shape standards across offensive security, defensive analytics, and AI-driven cybersecurity.

Education

• USP - Universidade de São Paulo

  Bachelor's degree, Computational and Applied Mathematics

• Uninove - Universidade Nove de Julho

  Doctor of Philosophy - PhD (Summa Cum Laude), Informatics

• Universidade Presbiteriana Mackenzie

  Master of Science - MS, Electrical and Electronics Engineering

  2007 — 2008