Conversations are logged for quality and AI security research.
Security Awareness Training·Human Risk Assessment·Workshops & Speaking·Cloud & AI Security
>_ FEATURED PROJECTS
Password Validator
DONE[01]
Weak passwords fail quietly until a breach makes them loud. This tool scores any password against breach databases, crack-time estimates, and security policy rules, in a CLI or a Streamlit web UI. Users see exactly why a password fails and what fixing it buys them.
Cloud dictation ships your voice to someone else's servers. RootVox is push-to-talk AI dictation for macOS built on whisper.cpp and Ollama: hold a key, speak, and cleaned-up text lands at your cursor. Fully local, zero cloud, zero telemetry.
Cloud text-to-speech ships your study files to someone else's servers and bills you for the privilege. PsyonVox loads a PDF, Word doc, PowerPoint, EPUB, or Markdown file and reads it aloud with the Kokoro-82M voice model running entirely in your browser: 27 voices, follow-along highlighting, and MP3 export, with your files never leaving the device. A companion Python script batch-converts whole folders to audio locally.
Cert practice apps grade you and move on; the miss is the moment worth teaching, and it gets skipped. CertRx captures your reasoning and uses AI to diagnose the specific misconception before revealing the answer. 11 live cert banks, 2,550 original questions, free with no account and no tracking.
AI workloads on AWS ship faster than the detections that watch them. CloudMind Sentinel attacks an Amazon Bedrock inference pipeline, then engineers real-time detection with CloudWatch, Lambda, and MITRE ATLAS. Attack and defense documented as one build.
Cloud & AI Security
RHEL Hardening
IN PROGRESS[06]
A default Linux install is not a compliant one. This project hardens RHEL against DISA STIGs with HIPAA safeguards mapped in, documented so the process is repeatable. Each control is verified, not just applied.
Foundations
Vulnerable Chat Bot
IN PROGRESS[07]
You cannot defend LLM apps you have never broken. This intentionally vulnerable chat application gives trainees a safe target for prompt injection and exploitation practice. Built to teach attacks so teams can recognize them in production.