Experience
Research Fellow
Jan 2025 – PresentMIT Senseable City Lab Rio · Rio de Janeiro, Brazil
There from the first week of the Rio branch. I work on morphometrics of informal urban form: describing favela fabric from terrain models and building footprints so that wind and sunlight simulation has something defensible to run on. Tech: Python, GeoPandas, point-cloud tooling.
Computer Vision and MLOps Intern (first intern)
Mar 2024 – Aug 2024Roboat (MIT Spinoff) · Amsterdam, Netherlands
Brought camera-based perception to a stack that until then ran on LiDAR: detection and multi-object tracking for autonomous canal vessels on Jetson Orin, and the MLOps around it, with DVC versioning over the on-water recordings the models were trained on. Tech: PyTorch, ROS 2, Docker, DVC.
Visiting Research Fellow
Oct 2023 – Feb 2024IRI (CSIC-UPC) · Barcelona, Spain
Built PyTorch fusion model combining neuromorphic event cameras with visual-inertial odometry for 6-DOF drone state estimation during aggressive maneuvers (>5g). Tech: PyTorch, ROS, C++.
Research Assistant
Jun 2022 – Jun 2023VITA Lab, EPFL · Lausanne, Switzerland
Built and open-sourced a full ROS autonomy stack for mobile robot navigation in crowded environments. Real-time detection, multi-object tracking, 3D pose estimation, and trajectory prediction on Segway Loomo. Tech: ROS, Python, C++, OpenCV.
Education
Master's degree (M.Sc), Robotics
EPFL
Thesis on urban tree canopy from aerial LiDAR, with Senseable City Lab, MIT and the AMS Institute.
Exchange (B.Sc 3rd year), Embedded Systems
KTH Royal Institute of Technology
Bachelor's degree (B.Sc), Microengineering
EPFL
Technical Skills
Robotics
ROS/ROS2, SLAM, path planning, sensor fusion, Jetson/edge deployment
ML / AI
PyTorch, scikit-learn, reinforcement learning
Perception
LiDAR processing, point clouds, stereo vision, event cameras, 3D reconstruction
Dev
Python, C++, Git, Docker, CI/CD, DVC, MLOps
Tools
OpenFOAM CFD, Blender, GeoPandas
| Robotics | ROS/ROS2, SLAM, path planning, sensor fusion, Jetson/edge deployment |
| ML / AI | PyTorch, scikit-learn, reinforcement learning |
| Perception | LiDAR processing, point clouds, stereo vision, event cameras, 3D reconstruction |
| Dev | Python, C++, Git, Docker, CI/CD, DVC, MLOps |
| Tools | OpenFOAM CFD, Blender, GeoPandas |
Languages
Open Source
Contributions and personal projects on GitHub: github.com/thrmnn