Member of Technical Staff

The Role

Lucid builds verifiable compute infrastructure: systems that produce cryptographic evidence about what AI hardware is doing, where it's doing it, and whether the computation matched what was claimed.

Where We're Hiring

Applied cryptography. Attestation protocols, proof systems, and the evidence formats that make a claim about compute independently checkable. You've built cryptographic systems that shipped, and you can reason about threat models as fluently as about primitives.

Hardware and silicon security. Roots of trust, TEEs and secure enclaves, and attestation on modern accelerators, including direct technical work with silicon vendors. You know GPU architecture and how logical tensors map to physical hardware.

GPU, compilers, and ML systems. Static analysis of kernels to extract memory access patterns and computational structure, and deriving what a workload should do from its definition. CUDA, Triton, CUTLASS, PTX/SASS, MLIR/LLVM.

Distributed systems and infrastructure. The platform that collects, stores, and serves verification evidence at cluster scale, plus the control plane around it. You've built distributed systems where correctness and availability both actually mattered.

Networking and measurement. Network taps, timing and latency measurement, and the inference problems underneath geolocating compute. You're comfortable with noisy physical-world signals and adversaries who know you're measuring.

Trusted environments and security engineering. Building and defending hardened compute environments, including physical and supply-chain security. Prior work in classified, regulated, or high-assurance settings is directly relevant.

If your specialty is adjacent to these but not on the list, apply anyway and tell us where you'd fit.

What Every Track Requires

• Real depth in your area. Years of it, or unusual velocity that substitutes for years. We'd rather have the person who's the local expert on one thing than four people who've touched everything.

• Strong systems programming. Rust, C, C++, or Go, with the correctness bar that comes with security-critical infrastructure.

• Ownership. You can take a problem from ambiguous question to shipped system without a spec or a manager checking in. At our size, the person who built it is also the person who explains it to customers, vendors, and regulators.
  

Bonus

• Formal methods or verification tooling

• Open-source contributions to infrastructure or security projects

• Experience at a company at a similar stage (sub-50 people, pre-Series A or Series A)

• Genuine interest in AI safety, governance, or responsible deployment
  

Note on level and scope

We're open to both early-career and experienced candidates. For the latter, we're open to higher compensation and more senior titles. We'll also consider exceptional candidates on a part-time basis.

Practicalities

This role can be based in San Francisco or London, or fully remote. The base salary range for this role is listed above. Compensation also includes equity.