Me

Daniel Layeghi

I am currently co-founder at Z2-labs, building Souida, an API-first discovery layer that turns natural-language book queries into structured real-time retrieval for independent booksellers. For an up-to-date CV, see here. Previously, I was a Postdoctoral Researcher and completed a PhD in Computer Science at the University of Edinburgh. My PhD thesis was titled "Optimal Control Theoretic Value Function Learning." Before academia, I was an early engineer at Automata Technologies and implemented and deployed a production-grade robotics stack from scratch.

Papers

  • Optimal Control via Combined Inference and Numerical Optimization. [ICRA 2022]
    Daniel Layeghi, Steve Tonneau, and Michael Mistry.

  • Neural Lyapunov and Optimal Control. [Under review 2023]
    Daniel Layeghi, Steve Tonneau, and Michael Mistry.

  • Learning Long-Horizon Robot Manipulation Skills via Privileged Action. [CoRL]
    Xiaofeng Mao, Yucheng XU, Zhaole Sun, Elle Miller, Daniel Layeghi, and Michael Mistry.