Initiatives

Building human-centric, data-aligned systems across robotics, infrastructure, and decentralized technologies.


🔹 Research-Driven Systems & Infrastructure
Building human-centric, data-aligned systems across robotics, infrastructure, and decentralized technologies.

Zero-Trust Data Infrastructure for Web3
🥇 Sui Foundation Hackathon
Built SIRIUS, a Git-like, zero-trust data layer that verifies and versions sensitive data before writing on-chain.
Enabled verifiable history, controlled updates, and auditability for sensitive datasets.

Human-in-the-Loop Robotics
🥉 AMD × Hugging Face Open Robotics Hackathon, STATION F
Developed an assistive robotic system that observes human tasks (assembly, surgery, maintenance), predicts the next required tool, and delivers it just-in-time.
Focused on robustness under distribution shift by separating intent (VLMs) from execution (imitation-learned skills): scaling better, generalizing further, and failing safer than end-to-end VLA policies.

AI Decision Systems for Healthcare & Climate
Designed AI tools for decision support under uncertainty, not full automation.


🔹 Learning Ecosystems & Community Leadership
Designing environments where people, constraints, and expertise collide to accelerate learning and execution.

Fellow, Unaite
Selected as a fellow in a national ecosystem uniting the top 7 French universities.
Mission: connect top students, researchers, and entrepreneurs through curated events, deep technical exchanges, and long-term collaboration.

AI & Research Lead, DeVinci Blockchain
Mentored 100+ students through workshops and applied research projects.
Designed the AI curriculum and facilitated exchanges with academic labs in decentralization research.

Organizer & Jury, IBM × De Vinci Hackathon
Co-designed a 48-hour sprint with IBM researchers and solution architects.
200+ participants tackling real-world industry problems under production constraints.

Co-Organizer, Shipping Coffee
Weekly build sessions where students and engineers prototype and ship complete systems in one evening.
Emphasis on fast feedback, execution, and learning in public.