Fraud & Security Analyst in the making. I build open-source tools that stop scams, theft and data exposure before they happen.
A decade in digital marketing and e-commerce gave me an insider's view of how fraud actually works: fake storefronts, social engineering, phishing and paid-ad abuse. Now I build the defenses.
Open to: fraud, SOC and security analyst roles · Remote or Southern California · Relocating to New York in 2027
7 security tools shipped · 110+ automated tests · real-world phishing takedown reported · incident response write-ups
Following my progress? Every lab, write-up and lesson is logged in cybersecurity-journey →
| Project | What it does |
|---|---|
| aegis-guardian | 🆕 A free, bilingual (English/Spanish) anti-scam agent: paste a suspicious text or screenshot and it explains the red flags and what to do next. Consent-first by design, so it can never become the surveillance tool it guards against. Bilingual scam checker live in the browser; AI layer powered by Meta's Muse Spark. Try it → |
| ai-guardian | Discovers the AI agents on a machine (apps, coding agents, browser extensions, MCP plugins) and flags what they can reach: plain-text API keys, SSH keys, disabled permission prompts. Enterprise-style "shadow AI" visibility, free and local. Project page → |
| php-guardian | Inventories every PHP install and flags versions exposed to php2shell, the critical flaw CISA warned about in Oct 2026. Shipped within a day of the alert; caught a real unpatched server in testing. JSON output for SIEMs. Project page → |
| wifi-guardian | Spots fake Wi-Fi before you join it: evil twins, look-alike network names, "FREE WiFi" bait and router swaps (ARP spoofing). Built for small businesses and guest networks without enterprise budgets. See the demo → |
| vpn-guardian | Verifies every 30 seconds that a VPN is actually protecting you, alerting laptop and phone the moment it drops or leaks your real IP, with a local AI that explains each alert in plain English. Project page → |
| beat-guard | Fingerprints producers' beats, timestamps them on Bitcoin and detects them inside other songs, even beneath vocals or sped up. Generates an evidence report. Project page → |
| tx-guard | Checks crypto transactions before you sign: matches 2,500+ known scam addresses, catches wallet-draining approvals and simulates the transaction to show exactly what you'd lose. Try it live → |
- Exposure review: AI coding-session links in public git history: threat model, secret scanning across eight repositories, false-positive triage and pre-commit controls
- Incident response: suspicious activity on a personal Google account: triage, credential rotation, session audit and persistence checks
Security: threat detection · incident response · fraud & scam analysis · threat modeling · secret scanning · network & wireless security · AI / LLM security Build: Python · Bash · Swift · Git · local AI (Ollama) · Ethereum · OSINT
- Building Aegis Guardian, a bilingual AI scam shield powered by Meta's Muse Spark
- Shipping the Guardian series hands-on: open-source security tools, each tested and documented
- Logging the work in my cybersecurity journey
Resume · LinkedIn · SoundCloud