PhD Candidate in AI & Robotics at IDSIA USI-SUPSI.
I work on embodied AI and robot learning, with a focus on connecting reasoning, perception, and physical interaction.
My research includes:
- knowledge-grounded LLM and RAG systems for robot planning;
- multimodal robot learning from vision, force, and contextual signals;
- diffusion policies and compliant manipulation;
- adaptive human–robot interaction and wearable robotics.
- ConceptBot — Knowledge-graph-grounded commonsense reasoning for robot task planning.
- Graph-Based RAG for Robot Planning — Provenance-aware retrieval with local LLMs and persistent knowledge graphs.
- Context-Aware Exoskeleton Assistance — Gaze, vision, and adaptive control for wearable robotic assistance.
- Learning Compliance with Diffusion Policies — Learning motion and compliance schedules for contact-rich manipulation.
Python · PyTorch · ROS 2 · Docker · LLM/RAG · Knowledge Graphs · Diffusion Policies · Computer Vision · Force / Compliance Control