From alert to action with explainable controls.
Storyline
How I think when the work has to survive real pressure.
Less "have you met Ted," more "have you met the threat model."
Make telemetry useful in motion, not just present in dashboards.
At Cantor Fitzgerald and NYU, I’ve focused on evidence paths that help responders decide faster: cloud logs, identity drift, threat intel, and the context needed to trust a next step.
- Evidence over volume
- Context over raw events
- Trustworthy signal over cosmetic coverage
"I want to live in a world where people can trust the stuff they use."
Automate policy and judgment, not just button clicks.
I build playbooks that encode standards: progressive blocks, recurrence-aware containment, quarantine review, and trusted-IP lifecycle management across multi-cloud infrastructure.
- Containment that can be defended later
- Playbooks that preserve operator trust
- Fewer loops that waste analyst attention
Stay technical enough to understand how defenses actually fail.
My background spans exploit solver scripts, ECU cryptography hardening, release-signing workflows, and post-quantum migration prototypes built for longer-horizon threats.
- Offense informs defense
- Crypto work informs architecture
- Attack-path thinking keeps controls honest
Push security teams toward governed, AI-native workflows.
The direction is clear: autonomous but supervised systems for detections, internal tooling catalogs, email triage, and operational reasoning that scale with confidence.
- Agentic triage with strong guardrails
- Better internal tooling discoverability
- Faster response without black-box drift