I build the layer between algorithms and actuators: robotics planning at Siemens, an autonomous industrial maintenance robot, vision models past 98% accuracy, LLM workflows in production. Open to AI and robotics engineering roles.
A few projects I'm proud of. Hover or tap any panel to read it.
Plans how a product is assembled by computing how it comes apart: automated sequence generation and pruning, path-finding with ML-RRT, D* and reinforcement learning, and AI-generated instructions exported to VR in Siemens NX. A Fourier-transform method estimates solution density across the inverse-kinematics configuration space, cutting placement computation time sharply.
LLM- and embedding-based workflows that harmonize the extracted diagram data: semantic deduplication, similarity search and clustering over thousands of components into clean, queryable groups.
The full software stack for an autonomous mobile robot working industrial floors: concurrent real-time systems in C++ and Python, 2D occupancy-grid SLAM, and sensor fusion under strict time-synchronization constraints. Coverage path-planning handles navigation and obstacle avoidance; a React and Node.js interface streams live maps over WebSockets for monitoring and remote teleoperation.
A pick-and-place arm driven end to end: real-time OpenCV detection with fiducial-based positioning, geometric inverse kinematics for a 4-DoF arm and wrist, and ESP32 firmware in C++ exposing a Wi-Fi socket server with state acknowledgments and over-the-air updates. A single Python loop closes camera capture, planning, communication and execution.
A custom-architecture YOLO model and computer-vision system that turns electrical single-line diagrams into structured digital data at over 98% detection accuracy, paired with a React tool stakeholders use to curate, cleanse, review and visualize the results.
Where I've worked and studied, the honors I've earned, and the tools I work with.
Based in Estado de Mexico, MX · +52 55 5474 9070