Junior Security Engineer · Cyvatar

Noah Cook

I build tooling that turns repetitive security analysis into one reproducible command: malware triage, vendor API reporting, and detection work.

About

I turn repetitive security analysis into tooling that scales.

I'm a junior security engineer at Cyvatar, working remotely on managed security operations for client organizations.

What I work on

  • Automating security reporting from vendor APIs like Tenable, SentinelOne, and Cloudflare
  • Turning scattered telemetry into reporting clients can act on
  • Putting AI systems like Claude to work on repetitive analysis

Most of my projects start as a problem I got tired of doing by hand, so I automate it.

Role
Junior Security Engineer, Cyvatar
Focus
Security automation and AI
Certifications
  • CompTIA Security+
  • AWS Certified Cloud Practitioner
Degree
B.S. Information Technology, FIU

Selected work

Tools I built to take repetitive analysis off my plate, plus a couple built for fun.

MAAT

Featured

Malware analysis automation tool

A command-line pipeline that triages suspicious files without detonating them. MAAT fingerprints a file by its magic bytes, hashes it in a single stream, and checks reputation through hash-only VirusTotal lookups, so the sample never leaves your machine. Analyzers dispatch by file type, findings feed an additive, explainable score, and the full run renders to a self-contained HTML report.

  • Python
  • YARA
  • VirusTotal API
  • oletools
  • pefile
  • pikepdf
  • Jinja2
MAAT report for jsdemo.pdf

Turns a manual, tool-hopping triage workflow into one reproducible command.

Marketplace Deal Finder

Personal project

AI-assisted deal scanner for Facebook Marketplace

A command-line pipeline that scrapes new listings in the location and category of your choice, uses the Anthropic API to identify the exact product from each listing's title, description and photos, and prices it against Google Shopping. The hard part is a retail price you can trust: results are stripped of used, resale and accessory listings, checked against the exact model by a second Anthropic API pass, and reduced to a median across stores. An item without a defensible price is dropped rather than guessed. The best discounts can optionally notify you in Slack, and SQLite remembers seen listings and cached searches so repeat runs cost less.

  • Python
  • Anthropic API
  • Apify
  • SerpAPI
  • SQLite
  • Slack

Scans 100 listings for about $2, and every run reports what it cost.

Heartwood Incremental

Personal project

Pixel-art incremental game

A browser incremental game where you grow a tree through escalating forms, funneling resources into three nested prestige-reset layers. The real work was systems design and balance: architecting a layered economy where every bonus multiplies through a single chokepoint, then tuning the pacing curve with a standalone simulator before any number reached the game. Built with AI-assisted development and shipped as dependency-free static files that redeploy on every push.

  • JavaScript
  • Canvas
  • HTML
  • CSS
  • Cloudflare
Heartwood gameplay capture

A complete product designed, balanced, and shipped end to end.

Skills and tools

The platforms I work in day to day, and what I build with.

Endpoint and email
CrowdStrike EDR · SentinelOne · Sophos · Endpoint Central · Mimecast · KnowBe4
Cloud and identity
Azure · AWS · Cloudflare · Microsoft Entra · DigiCert
Vulnerability management
Tenable · Vonahi
Automation and analysis
Python · YARA · Power BI · DAX · Power Query

Contact

I'm open to hearing about security roles where I can build tooling and put AI to work on real analysis. Email is the fastest way to reach me.