Photonics AI Engineer - Component Inverse Design

Xanadu•,

21d•$160,000 - $210,000

About The Position

Xanadu is seeking a Photonics AI Engineer to own component-level design on a fast-paced path-finding team. The role designs passive and active components for photonic quantum hardware, aiming for sub-milli-dB loss and broadband, fabrication-robust performance. The engineer will use and build agentic AI automation and ML/AI-driven optimization, treating design tooling as a first-class deliverable. The position takes component concepts through to verified building blocks ready for tapeout. It involves high-throughput electromagnetic simulation and optimization, designing for manufacturability, and collaborating with layout, fabrication, and test engineers to compare designs with measured wafer data. The engineer will also contribute reusable design and optimization infrastructure to the team’s codebase. The team primarily uses Python and its scientific ecosystem, commercial and open-source electromagnetic solvers, cloud-based high-throughput simulation, GitHub, Docker, and agentic AI tools. Proficiency in these technologies is essential.

Requirements

Fluency with modern AI tools in day-to-day work, such as coding agents, LLM-driven automation, or agentic pipelines, and a track record of building tools.
Participation in multiple tapeouts covering design, layout, and post-silicon analysis, with components that were fabricated and measured.
Hands-on experience applying ML/AI to a technical problem in any domain.
Enthusiasm for quickly adopting new AI capabilities and reshaping workflows around them.
BSc in Engineering, Physics, or a related field.
Strong knowledge of integrated photonic component design, including device loss sources and their relationship to fabrication processes.
Experience with electromagnetic simulation, including FDTD, EME, FEM, or mode solvers, and understanding of solver-fidelity and computational-cost trade-offs.
Experience with automated design optimization rather than manual parameter sweeps.
Strong Python skills, including reusable, tested, version-controlled code.
Comfort operating in a fast-paced environment with multiple competing priorities.
Proficiency with Python and its scientific ecosystem, commercial and open-source electromagnetic solvers, cloud-based high-throughput simulation, GitHub, Docker, and agentic AI tools.

Nice To Haves

MSc or PhD in Engineering, Physics, or a related field.
Experience building agentic systems or AI tooling for a technical domain.
Experience with ML/AI-driven or inverse design of physical systems.
Experience designing active components such as electro-optic modulators or photodiodes.
Experience with robust or yield-aware design under process variation.
Experience executing tapeouts with external photonic foundries.
Familiarity with quantum photonics or quantum computing hardware.

Responsibilities

Own component design from physical concept through to a verified building block ready for tapeout.
Apply ML/AI-driven inverse design and AI-assisted optimization to passive and active integrated optical components.
Run large simulation campaigns across FDTD, EME, FEM, and mode solvers, accounting for convergence, numerical error, and compute cost.
Establish agentic AI tooling and processes to launch simulation campaigns, triage results, and extract figures of merit.
Design for manufacturability so that simulated performance can be achieved in fabrication.
Work with layout, fabrication, and test engineers to close the loop from concept to measured wafer data.
Contribute reusable design and optimization infrastructure to the team’s codebase.