Research Engineer - AI Verification

Your work

As a Research Engineer, you'll help turn emerging ideas in AI verification into practical technical prototypes. We're seeing more interest in our prototypes than we currently have capacity to pursue, so additional engineering capacity will directly expand the mechanisms and approaches we're able to explore.

AI verification is still an open technical problem. Mechanisms need to be robust to attempts to evade them, privacy-preserving enough to be deployable, and auditable enough to earn the trust of governments and other stakeholders. Solving these problems can draw on ML engineering, cybersecurity, cryptography, hardware, and systems engineering.

Our current project involves distinguishing between inference and training workloads on GPUs. Future projects could include scaling zero-knowledge proofs of AI inference and developing privacy-preserving approaches to white-box evaluations.

In this role, you would

• Build prototypes supporting AI verification.
• Support technical communications.
• Collaborate with external experts.
• Work across: ML engineering, inference pipelines, algorithms, frontend demos, and cybersecurity.

• AI hardware engineering.
• Confidential computing.
• Cybersecurity instrumentation.
• Startup experience.
• Interest or experience working with Singaporean or Chinese stakeholders.

• Strong ML engineering fundamentals.
• Familiarity with AI verification, cybersecurity and/or confidential computing.
• Comfortable in ambiguous early-stage environments.

• This is a remote role that can be based in Singapore, the UK, or remotely.
• For candidates interested in relocating, SASH can provide visa sponsorship in the UK.
• Our team is globally distributed, so remote team members should be comfortable maintaining some working-hour overlap with colleagues across regions.

Compensation

Our compensation takes location, experience, and level into account. Salaries above the stated range may be available for exceptional candidates.

Benefits

• Competitive benefits and leave policies.

AI verification is still a young and highly talent-constrained field, while the need for credible mechanisms to support international coordination around advanced AI is becoming increasingly prominent, including in initiatives like Pacing the Frontier https://www.pacingthefrontier.com/?utm_source=chatgpt.com.

Joining SASH now means working on technical problems where the solution space is still open, with significant room to explore approaches, build prototypes, and influence what gets developed, what the team builds, and how this emerging field connects to real-world AI governance. Scenarios like AI 2040 https://ai-2040.com/?utm_source=chatgpt.com illustrate the role technical verification mechanisms could eventually play in making international AI agreements credible in practice.

SASH is a young, fast-moving organization where finding opportunities and making things happen is the norm. You'll have substantial autonomy to build around promising ideas rather than inherit a mature roadmap, while working with an international network of technical experts and policymakers.