David Ortiz / Razon RazonIn4K

Hi, I'm David Ortiz

I build practical web systems, automation workflows, and AI-security tools from DeKalb, IL.

My current work sits in two lanes:

HighEncode Learning: local-business websites, lead-flow automations, and operator tooling for small teams.
Prompt defense and AI security: prompt-injection scanning, rules-based guardrails, and evaluation workflows for AI-assisted systems.

I care about evidence-backed builds: working demos, readable code, documented tradeoffs, and checks that make the next iteration easier.

Shipping focused websites and automation systems for operators who need clearer intake, follow-up, and handoff flows.
Building and polishing Prompt Defenders as a prompt-injection scanner and rules engine.
Turning AI-security evaluation work into public-safe defensive artifacts: threat models, checklists, scoring rubrics, and reproducible toy cases.
Using multi-AI workflows carefully: source checks first, tool boundaries explicit, and public claims verified before publishing.

Privacy-first prompt-injection scanner with a versioned rule pack, severity scoring, API surface, CLI workflow, and public rule documentation.

Repo: https://github.com/RazonIn4K/prompt-defenders
Use case: pre-LLM request scanning, prompt-library review, and guardrail triage.

Personal portfolio and notebook for selected work, operating notes, and project experiments.

Commercial surface for local-business websites, practical automations, and AI tooling.

Full-stack job-board project using Next.js, Firebase, Stripe, TypeScript, and role-based dashboards.

Web and product: Next.js, React, TypeScript, Tailwind CSS, shadcn/ui, Vercel

Automation and data: Python, Bash, Playwright, APIs, Firebase, Firestore, Stripe

Security and AI workflows: prompt-injection testing, rule-based detection, threat modeling, MITRE ATT&CK/ATLAS mapping, Burp Suite, Nmap, ffuf

Operating style: source-first research, browser verification, tests where they matter, and written handoff notes.