# Jorge Cabello — Product Engineer

> Self-taught product engineer with 7+ years at startups. Frontend, native apps, AI tools. Based in Buenos Aires, Argentina. Open to senior frontend and full-stack roles (remote preferred).

## About

I taught myself to code in Venezuela and moved to Argentina in 2019. Started doing creative frontend work on Fiverr. GSAP animations, SVG stuff, the kind of sites that actually move.

Then I joined a startup as the second engineer, built the frontend from nothing, and stayed through Series B as the team grew to 30. After that, more startups. Always early, always building the thing that didn't exist yet. Right now I'm three years into a funded EdTech company where I went from frontend to leading the AI work.

React, TypeScript, and animation are my core. I build with Claude Code and Codex as daily collaborators. Custom tooling, multi-agent reviews, automated workflows. I also ship native apps in Swift on the side. Frontend is home.

## Experience

### Senior Software Engineer — Cometa (2022–Present)
Series A, $12M raised, K-12 SaaS. React, Next.js, TypeScript, Turborepo.

**Frontend & Architecture:**
- Set up the frontend monorepo from early on: Turborepo, React Query, shared configs. Three apps run on it today.
- Built most of the core product surfaces: payments, invoicing, school management, reporting, data tables. End-to-end feature ownership.
- Built the AI chat interface (Jarvis). Streaming responses, conversation history, feedback, accessibility. Used daily by every team at the company.
- Kept the platform stable and fast as it scaled. Deployment pipeline, error tracking, performance.

**AI Products:**
- Built the frontend for a support agent that handles 80% of incoming tickets. Streaming chat UI, feedback system, admin dashboards.
- Shipped the interfaces for RAG search and a SQL agent with per-school data isolation. Built the query UIs and dashboards that four non-technical teams use daily.
- Ran hackathons, lightning talks, and AI 101 sessions. Taught 50+ people across the company how to actually use these tools.

### Senior Full-Stack Developer — Flevo (2022)
Educational Fintech, Buenos Aires.
- Only frontend engineer. Built the checkout flow and school management dashboard from zero.
- Credit scoring UI that took loan processing from days to 24 hours. Over 1,000 students approved through the system.

### Full-Stack Developer — Karvi (2019–2022)
Seed to Series B, used-car e-commerce, 100k+ monthly users.
- Joined as engineer #2. Built the frontend from scratch: vehicle catalog, dealer dashboard, real-time inventory.
- Helped grow the team from 2 to 30 through Series B. Set the standards, onboarded new engineers, kept quality high while shipping fast.

### Creative Frontend Developer — dokstudio (2017–2019)
Freelance, Fiverr, international clients.
- SVG animations, GSAP transitions, interactive experiences. 50+ sites for clients across the world.
- This is where I started caring about craft. How things move, how they feel.

## Selected Projects

### Creative Dev (Freelance)
Two years of freelance creative frontend. GSAP animations, SVG effects, interactive marketing sites. Built for agencies and startups around the world.
Tags: GSAP, SVG, CSS, JavaScript
Link: https://codepen.io/dok

### Claudio (Web + Bot)
Telegram bot and web UI that lets me control Claude Code sessions from my phone. Streams output in real-time, manages tmux sessions, parses and formats everything nicely.
Tags: React, TypeScript, Bun, Telegram API

### AI Platform (Enterprise, Cometa)
The AI work at Cometa: a support agent that handles 80% of tickets, RAG search over company data, a SQL agent that respects per-school permissions, and dashboards non-technical teams actually use.
Tags: Python, RAG, pgvector, Next.js, AWS

### Zas (macOS)
macOS menu bar app that runs speech-to-text locally on Apple Silicon. No cloud, no waiting. Press a key, talk, it types. 25 languages, filler word removal, runs entirely offline.
Tags: Swift, SwiftUI, Apple ML, C++ Interop

### Voiz (iOS)
iOS journaling app where you talk instead of type. Local transcription, emotion tracking over time, AI-generated insights about your patterns.
Tags: SwiftUI, SwiftData, On-device ML, OpenRouter

## Stack

**Frontend:** React, Next.js, TypeScript, Tailwind, GSAP, SwiftUI
**Backend:** Node.js, Python, FastAPI, PostgreSQL, AWS, Docker
**AI / Tooling:** Claude Code, Codex CLI, RAG, pgvector, MCP, AI Agents
**Native:** Swift, SwiftData, Apple ML, Xcode, C++ Interop

## AI-Native Workflow

I treat AI tools the way most engineers treat their IDE. They're part of the infrastructure. These are custom Claude Code skills I built to automate real workflows:

- **/council** — Multi-model code review. Sends context to Codex for independent analysis, then compares perspectives. Catches blind spots neither model finds alone.
- **/sentry-fix** — End-to-end production error resolution. Finds the error in Sentry, traces it to the code, applies a fix, opens a PR, and monitors the deploy.
- **/recall** — Semantic search over 7,100+ files of past conversations and notes. Gives every new session full context from every previous one.
- **/finance** — Browser automation that logs into my bank, pulls transactions, categorizes spending, and updates financial tracking. Full Playwright pipeline.
- **/afip** — Navigates Argentina's tax portal to generate export invoices. Form filling, validation, submission. All through browser control.
- **/light** — Natural language control of Philips Hue lights. Mood presets, room awareness, bilingual commands. Hardware integration through Claude Code.

## Contact

- Email: jorgecbll@gmail.com
- GitHub: https://github.com/dokeet
- LinkedIn: https://www.linkedin.com/in/jorge-cabello-83a9b816b/
- CodePen: https://codepen.io/dok
- Location: Buenos Aires, Argentina