Post-Training AI Research Scientist: RLHF for Finance

Two Sigma; New York (NY); Hybrid; USD 165,000 - 300,000; Full time

Benefits offered by this job

Health insurance

401k match

Life & disability insurance

Gym access

Tuition reimbursement

Generous vacation & sick days

Hybrid work policy

Home office budget

Job summary

Two Sigma is advancing post-training with RLHF, DPO, and reward modeling to align LLMs with complex, multi-step financial workflows. You will define the research agenda, build scalable infrastructure, and guide evaluation frameworks for transforming quant research into production-grade AI capabilities.

The role combines training, fine-tuning, context management, and model evaluation, shaping both the post-training capability and the broader research direction of the team.

Qualifications

• BS or equivalent in Science, Technology, Engineering or Math (an MS is a plus).
• Minimum 1 year of experience required; 1-10 years preferred at a frontier AI lab (OpenAI, Anthropic, DeepMind, Meta FAIR, or equivalent).
• Shipped post-training systems in production: RLHF, DPO, RLAIF, or related methods.
• Deep understanding of distributed training infrastructure: multi-node GPU clusters, training stability, checkpointing.
• Track record managing large-scale compute: budgeting, experiment design, ablations.
• Publications or demonstrated expertise in alignment, preference learning, or reward modeling.
• Hands-on implementation skills: PyTorch/JAX, distributed frameworks (DeepSpeed, FSDP, etc.)

Responsibilities

• Lead post-training efforts for LLMs applied to financial time series and quantitative reasoning
• Design and execute RLHF, DPO, and related alignment methods at scale, including deployment of substantial compute budgets (O($100mm))
• Build infrastructure for preference data collection, reward modeling, and policy optimization on financial datasets
• Drive research agenda connecting post-training methods to quantitative finance applications
• Collaborate with quant researchers to define task distributions and evaluation frameworks
• Unblock production systems dependent on post-training capabilities

Skills

PyTorch/JAX

Distributed training

RLHF/DPO knowledge

LLM alignment research

Experiment design

Budgeting compute

Education

BS or equivalent in STEM

MS preferred

Tools

DeepSpeed

FSDP

Job description

Two Sigma is advancing post-training with RLHF, DPO, and reward modeling to align LLMs with complex, multi-step financial workflows. You will define the research agenda, build scalable infrastructure, and guide evaluation frameworks for transforming quant research into production-grade AI capabilities.

The role combines training, fine-tuning, context management, and model evaluation, shaping both the post-training capability and the broader research direction of the team.