GRC leader with 12+ years building compliance, risk, and security programs at high-growth tech companies and public-sector orgs. I don't just manage audits. I write the code that automates them.
I started in the trenches: racking servers, writing PowerShell scripts, managing VMware clusters. That sysadmin foundation taught me how infrastructure actually works, which turned out to be the thing most GRC people are missing.
From there I moved into security operations and eventually into GRC program management, leading risk programs and building a vendor assessment process covering 2,000+ vendors for the City of San Francisco (55 departments, 30,000 employees) before shifting to high-growth startups.
At Amplitude and Handshake, I found my groove building compliance programs that actually work. SOC 2, ISO 27001, and PCI DSS on the company side. SSPA, bank due diligence, and Mag 7 customer assessments on the other. I built the programs and simultaneously built the internal tooling to automate the boring parts. This is also where I learned firsthand how painful customer trust workflows are and became convinced that no human should ever manually fill out a security questionnaire again.
AI has been part of my toolkit since the early GPT-3.5 days, when I started using it to accelerate GRC workflows like policy drafting, risk analysis, and evidence mapping. That evolved into writing Python scripts with Copilot, then building full production applications with Claude. Today I use AI-assisted development to ship internal security tools on GCP, going from idea to deployed Cloud Run service in days, building tools that would otherwise never exist.
The throughline: I make security programs that don't depend on me being in the room. Automated evidence collection, self-service vendor reviews, codified risk processes. Programs that outlast the audit cycle.
End-to-end ownership of audit programs from scoping through delivery. Zero-exception track record across SOC 2, ISO 27001, and PCI DSS. Built automated evidence collection pipelines that cut audit prep time by 80%.
Build governance and risk programs from scratch: policy frameworks, risk registers, risk councils, and assessment methodologies. Experienced with FAIR quantitative risk and ISO 27005 qualitative approaches.
Design, build, and deploy production security applications using Claude Cowork. Full-stack apps on GCP Cloud Run that automate vendor risk, customer trust, and compliance workflows, shipping tools in days that would otherwise never exist.
Deployed self-service customer trust portals and AI-powered questionnaire response, reducing time spent on security questionnaires by 90%+. Built vendor risk programs assessing 2,000+ vendors with risk-based tiering and SLA-driven review processes.
Enterprise-scale security awareness and training programs with onboarding controls, phishing simulations, and completion enforcement. Build security culture through education, not just policy.
Deep infrastructure roots: VMware, Active Directory, Linux/Windows ops. PowerShell and Ansible automation. The foundation that makes everything else possible.
A scheduled job reads an identity provider’s configuration daily, compares it with settings approved in Git, stores the result as locked audit evidence, and posts to chat. Continuous monitoring built from a few small parts.
Automated vendor risk platform that syncs with procurement (Ramp) and project tracking (Linear). Tracks 162 vendors with automated risk evaluation, daily syncs, and Slack alerts.
A report-only agent that runs a daily standup over my open issues and scheduled agents. It reports what moved, what has gone quiet, what looks finished but was never closed, and which agents ran. It never changes anything.
.claude/skills folder and they work immediately. Built from real production workflows, anonymized for public use.A report-only template for a daily standup over your open issues and scheduled agents. Reports what moved, what has gone quiet, what looks finished but was never closed, and which agents ran. Placeholders for your own tracker, agent log, and schedule.
View skill →End-to-end automation for security risk management. From raw input to a fully populated risk register row and linked tracking issue. Supports structured assessments and quick intake from unstructured notes, chat threads, or pentest findings.
View skill →A reusable pattern for automating vulnerability management operations. Drop it into any Claude project and it handles reporting templates, triage routing logic, exception workflows, and program hygiene rules.
View skill →A screenshot proves what a setting said on one day. This post describes how to read a system’s configuration every day, compare it with settings a person approved in Git, keep the result as audit evidence, and alert on differences. Okta is the example, and build or buy is the lens.
Read post →Building programs with Claude changed how I work. This post collects what held up across the previous four posts and what remains open.
Read post →An agent that reads records carries one set of risks. An agent that changes them carries another. This post covers the design choices Claude and I use: report-only, propose-only, independent review before execution, and an iteration cap.
Read post →