Fullstack Developer · Python Automation · IPN Student
Building educational technology, IoT & trading tools that solve real problems
| Education | Focus | Location |
|---|---|---|
| CECyT No. 3 "Estanislao Ramírez Ruiz" — IPN | Fullstack (TS/React/Next.js) + Python Automation + IoT | EDOMEX, México |
- Currently building: CECyT-3-Lab — Interactive virtual lab for Physics, Chemistry & English (Next.js 16, React 19, TypeScript, Tailwind 4)
- IoT & AI: SCARI — Smart waste-sorting bin with ESP32 + Groq AI vision
- Automation enthusiast: trading-bot — Disciplined DCA bot for Binance with 11 failed strategy attempts documented
- Learning: Advanced TypeScript patterns, System Design, Rust basics
- Philosophy: Fail productively → Learn systematically → Build reliably
- Open to: Junior/Intern roles, freelance projects, open source collaboration
| Project | Description | Stack | Status |
|---|---|---|---|
| SCARI | Smart waste-sorting bin with AI vision — ESP32-CAM captures, Groq AI classifies organic vs inorganic, NEMA 17 sorts. Fully documented & tested | ESP32-S3 ESP32-CAM Arduino Groq API AI Vision |
Completed |
| CECyT-3-Lab | Virtual lab for Physics, Chemistry & English with productive failure pedagogy | Next.js 16 React 19 TypeScript Tailwind 4 Prisma 3Dmol |
Live cecyt3-lab.vercel.app |
| mi-portafolio | Personal portfolio — software + hardware & IoT projects | HTML CSS JS |
Live mi-portafolio-phi-six.vercel.app |
| trading-bot | Disciplined DCA bot for Binance — 11 strategies tested, documented pivot to automation | Python CCXT Pandas Binance API |
Testnet validated 16+ trades |
- Smart waste-sorting bin that classifies trash into organic vs inorganic using AI vision
- ESP32-CAM captures frames → ESP32-S3 classifies via Groq API (
llama-4-scout-17b-16e-instruct) - NEMA 17 stepper motor sorts waste into compartments with trapezoidal acceleration profile
- Native RGB LED (WS2812 on GPIO48) provides instant visual feedback
- Async web portal with live dashboard, accuracy gauge, history log, and REST API
- Fully documented — architecture, hardware wiring, API config, troubleshooting guide
"The user just drops the trash — SCARI detects, classifies, and sorts it all by itself."
- 10 interactive modules across 3 subjects (Physics, Chemistry, English)
- Pedagogical sequence: Productive Failure → Simulator → Worked Example
- 38 tests (Vitest + Playwright) · WCAG AA · Touch-friendly (44px targets)
- Design system with subject-based color tokens (Emerald/Fuchsia/Amber)
- Documentation: Project Charter, Roadmap, WBS, Design System, Pilot Report with statistical analysis
- Scientific approach: 11 strategies tested (SMA, RSI, Arbitrage) → all failed out-of-sample
- Documented lessons: Survivor bias, overfitting, real costs (fees+slippage), 30-trade minimum rule
- Robust DCA: Non-blocking file locks, multi-currency, testnet/mainnet toggle, CSV logging
- GitHub Actions configured (blocked by Binance IP restrictions — documented workaround)
"Code is read more than written. Make it readable, testable, and honest."
Star my repos if you find them useful!
This profile README is generated/updated via GitHub Actions
- Stats cards:
github-readme-stats(Vercel) - Streak:
github-readme-streak-stats(Heroku) - Snake animation:
github-contribution-grid-snake(GitHub Actions) - Visitor count:
komarev.com/ghpvc