Day work
Lead Security Architect work across banking, insurance, and AI systems. Architecture reviews, threat modelling, and the awkward seams where the three overlap.
cat ./now
Now.
What I'm focused on right now. Updated roughly monthly so it stays current.
> Last updated
Writing
Long-form practitioner essays on the security of AI systems and the practice of cybersecurity architecture more broadly. New post most months at /blog/.
Research thread
Five-zone threat modelling for production agentic systems, translating method into a worksheet engineering teams can use without me in the room.
Open source
prose-scrub, detect and rewrite LLM-style prose. Plus a chatbot-app demo of LLM red-teaming with PromptFoo. Both at github.com/Antak108.
Speaking
Four to six talks a year. Most recent batch focused on red teaming generative AI systems and MAESTRO-aligned ML security. Topics + venues at /speaking/.
Reading
Papers on agentic AI safety and adversarial robustness. Practitioner writing from people doing security on AI systems at scale.