Verification Engineer; Fully Remote

• Own the threat model: enumerate what malicious or careless workers can do across pre-training, post-training, and inference.
• Design and calibrate statistical tests with stated error rates, tuned via rigorous benchmarks.
• Ship the verifier: build and run the verification service in the inference path and live with its mistakes.

What We're Looking For:

• Shipped or published verification systems with calibrated statistical decision making.
• Deep statistical depth to design experiments and defend error rates.
• Technical knowledge of verifying untrusted compute, from statistical testing to cryptographic proofs and trusted hardware.

Nice to Have:

• Familiarity with large-scale pre-training and RL post-training.
• Familiarity with decentralized ML security and adversarial threat models.
• Experience at proprietary, open-weight, or open-source AI labs.
• Equity-heavy package with significant ownership for key technical contributors.
• Remote-first culture with flexible work across global timezones.
• Optional full visa sponsorship and relocation support to Australia or the US.

Pluralis Research is building Protocol Learning, a decentralized approach to training and serving large models on consumer-grade devices. They are a small, world-class team of ML researchers backed by Union Square Ventures, with a remote-first culture and an ideological commitment to trustless, sovereign AI.