A focused team building infrastructure and methods that make open models systematically better after initial training.
Member of Technical Staff — Post-Training Research
Referral bonus eligible. Know someone who would be great for this role? Refer them to a Fluency Digital recruiter. If they're placed, you may be eligible for a referral bonus of up to $5,000 (terms apply). Refer someone for this role
In your first year, you will design and run experiments that meaningfully shift the behavior of large language models — improving reasoning depth, instruction fidelity, and safe refusal boundaries. You will own end-to-end research loops: forming hypotheses about where post-training falls short, scripting large-scale evaluations in PyTorch, and shipping improvements that compound across model versions. This is not a lab notebook role; your work will land directly in production fine-tuning pipelines that serve real downstream users. The team is small enough that your fingerprints will be unmistakable, and the research agenda is still being shaped.
What they're looking for
- 2–6 years of hands-on experience training or fine-tuning large neural networks, with deep fluency in Python and PyTorch
- Track record of empirical research in areas such as RLHF, DPO, constitutional AI, red-teaming, or related alignment and post-training techniques
- Comfort designing rigorous experiments at scale: building data pipelines, instrumenting metrics, and making tradeoffs between cost, latency, and model quality
- Strong communication instincts — you can defend experimental design choices and distill failure modes into clear next steps without relying on jargon
- Bias toward action in an environment where priorities shift with fresh experimental evidence; you would rather run the small pilot tonight than over-plan for a perfect run next week