Robotics Engineering @ Tec de Monterrey, Campus Guadalajara
Telemetry and power systems for SAE teams · autonomy · AI orchestration
- 🏎️ MadRams (Baja SAE) and Quantum Speed Racing (Electrathon MX): building the vehicle telemetry and EV power systems from scratch
- ⚡ Silca Elyos (Shell Eco-marathon): energy optimization on the team's existing car, including pulse-and-coast strategy and FOC tuning. I also lead the team's new autonomous-car (ADC) effort
- 🤖 OmniSim: follow-up PRs on obstacle-avoidance tests and procedural maps, building on my merged Nav2 contribution
| Project | What it is | Stack |
|---|---|---|
| JTCS · demo | Edge-AI adaptive traffic signals. Guadalahacks 2026 | YOLOv11 · Python CV · React · Node WS · RPi |
| TelemetryStack | Reusable telemetry base I built from scratch, used by 2 SAE teams so far | ESP32 · FreeRTOS · LoRa 915 MHz · MQTT · InfluxDB · Grafana |
| MadRams Baja car | Vehicle comms and telemetry. The team placed 4th in Endurance at Baja SAE Oregon | ESP32 · MQTT/Mosquitto · LTE/WiFi · u-blox GPS · SX1262 |
| OmniSim Nav2 | Packaged Nav2 stack for Husky diff-drive. PR #12 merged and headlined in release v8.1.17 | ROS 2 · Nav2 · MuJoCo · CycloneDDS |
| Banky · demo | Finalist, Banorte challenge at HackMTY 2026. Agent-built financial UI; I built the whole backend and harness | LLM · MCP · A2UI · Anthropic / Gemini |
| NeuroBeat | EEG signal-processing project | EEG · Python |
| Campus Map · demo | Pedestrian routing map for Tec GDL | GeoJSON · JS |
| MadRams Course Library · demo | Telemetry and CAD onboarding courses for the Baja team, with sessions mapped to the real car's subsystems | Web · SolidWorks · Three.js |
🧠 AI orchestration and Claude Code skills
- LaTeX Studio: local IDE that turns PDFs into APA-7 LaTeX. I built the pipeline (FastAPI, PyMuPDF extraction, Monaco, pdflatex); an LLM generates the
.tex - claude-unlimited: multi-agent orchestrator with per-role fallback over a self-hosted LLM gateway
- websight: headless-browser visual QA skill
- exploded-view: web-native 3D presentation skill (model-viewer, Three.js, glTF/Draco)
- watermarks-remover: Claude Code plugin integration for guillaumemeyer/watermarks-remover. The engine is upstream's work
🦾 Earlier robotics (high school)
- Robot Interprepas: ROS 2 Humble diff-drive robot with Nav2, LiDAR and Arduino
- AgroBot · site: raspberry-harvesting prototype with a 6-DOF arm, YOLOv8n and inverse kinematics
Also: LoRa/SX1262 · CAN/TWAI · FOC/BLDC control · GPS NMEA · LVGL · BLE reverse-engineering (bleak) · YOLO · Ollama and local LLMs · MCP
| Language | Depth | Used for |
|---|---|---|
| C++ | Core | Embedded control and firmware: FOC/BLDC tuning, FreeRTOS on ESP32, sensor drivers |
| Python | Strong | Computer vision, ROS 2 nodes, data pipelines, FastAPI backends |
| TypeScript / JS | Practical, AI-assisted | Web demos, dashboards and course sites that present the engineering |
GitHub's language stats count lines, so the web demos outweigh the firmware. This table is the honest version.
I like controlling tools below their intended abstraction: BLE protocols, file formats, CLIs as engines.