AI & Security Engineering
I build AI. I also break it.
Six years across AI and security engineering, the last two leading a team of ten. Most days I'm building agentic LLM pipelines with LangGraph (Python or TypeScript) and shipping them on AWS, Azure or GCP with the MLOps wiring that goes with it: experiment tracking, model versioning, observability across the lifecycle. The other half of the job: red-teaming LLM systems against the OWASP LLM Top 10 with tools such as Garak or Promptfoo, and wiring DevSecOps into the CI/CD.
Engineering & research
Featured writing
- llm-security 18 min read
Building an Agent That Can't Be Hijacked: The Design Rules
Eight frameworks tell you how to secure an agentic system and none of them are substitutes for each other. What they agree on, where they actively contradict each other, and the four properties I'm going to enforce in code in Part 2.
Read article - llm-security 26 min read
I Built the Defense, Then Attacked It: Numbers From a Secure Triage Agent
A capability-tracking interpreter for a SOC triage agent, a 20-case injection suite aimed at it, and 420 runs across three models. Exfiltration and disclosure go to zero and stay there regardless of model strength. Evasion tracks model capability, and one of my four policies turned out to be pure friction. Includes the LangGraph port, and why the framework's flagship agent pattern breaks the design.
Read article - llm-security 23 min read
Why a Classifier Won't Save You From Prompt Injection
Four documented ways to evade Prompt Guard-style detectors, and two underlying results — a cryptographic impossibility barrier and a measured security-fidelity tradeoff — that explain why this isn't a problem you fix with more training data.
Read article - ai 8 min read
Building RedTeam MCP: An AI-Powered Penetration Testing Assistant
A practical guide to building an MCP server that enables AI assistants to orchestrate offensive security tools for penetration testing, with proper safety guardrails
Read article - langchain 3 min read
Building LangChain Tools and Agents: From Zero to SOAR Assistant
Learn how to create LangChain tools from scratch and build a simple SOAR (Security Orchestration, Automation and Response) agent. This hands-on tutorial covers the fundamentals of tool development and agent creation for security automation.
Read article
Latest posts
All posts →- rag 5 min read
Building a RAG System for Cybersecurity Compliance: A Simple POC with LangChain v1+
A step-by-step walkthrough of building a basic RAG system with LangChain v1+ to query compliance regulations. This is an unoptimized POC for learning purposes, not production-ready.
Read article - Elastic ML 12 min read
Engineering Security ML with Elastic – Part 4: Production Pipelines with Dagster & MLflow
Transitioning from exploratory notebooks to reliable, repeatable anomaly detection pipelines using Dagster and lightweight MLflow-based model management.
Read article - ml-cyber 20 min read
Engineering Security ML with Elastic – Part 3: Benchmarking Unsupervised Models
How to Compare Anomaly Detection Models Without Labels
Read article - langgraph 5 min read
LangGraph Agent Architectures and Patterns: A Professional Guide
A comprehensive theoretical guide on agent architectures and patterns used professionally in AI projects. From simple workflows to hierarchical multi-agent systems, with real-world use cases for each pattern.
Read article - elasticsearch 16 min read
Engineering Security ML with Elastic – Part 2: Unsupervised Anomaly Detection
Learning Normal Authentication Behavior with Autoencoders
Read article - elasticsearch 11 min read
Engineering Security ML with Elastic – Part 1: From Logs to Features
From Windows Event Logs to Behavioral Features: Preparing Elasticsearch Data for Anomaly Detection
Read article