AI research scientist

WRITER•San Francisco, CA

1d•$168,000 - $233,000•Hybrid

About The Position

WRITER is hiring an AI Research Scientist to advance research in large language models, agentic reasoning, and system-level capabilities for enterprise AI. The role focuses on research that both advances the field and translates directly into products used by hundreds of thousands of people. The scientist will contribute to work on post-training, planning, multi-step reasoning, and agentic workflows, supported by the company’s research infrastructure and cross-functional teams. The position is hybrid and based at WRITER’s San Francisco, Seattle, or New York City hub. It reports to the VP of AI Research. WRITER is open to hiring at multiple levels, including senior applicants, with compensation scaled to experience, expertise, and role scope.

Requirements

At least 4 years of hands-on machine learning research experience, which may include time spent obtaining an advanced degree.
Deep expertise in large language model pre-training and post-training, including experience training models at scale, debugging distributed jobs, and shipping measurable improvements.
A Ph.D. in Computer Science, Machine Learning, NLP, or a related field, or equivalent demonstrated research experience with a strong portfolio of independent published work.
Expert knowledge of post-training methods including SFT, RLHF, RLAIF, DPO, GRPO, and related alignment and reasoning techniques, with experience applying them to production-grade systems.
Strong command of Python and PyTorch or JAX, and the engineering ability to build and scale training pipelines, evaluation infrastructure, and data synthesis workflows.
At least one publication at a competitive ML/AI venue such as NeurIPS, ICLR, ICML, ACL, EMNLP, or equivalent.
Hands-on experience designing or evaluating agentic systems that plan, reason through multi-step tasks, use tools, and recover from errors.
Ability to collaborate across research, engineering, and product teams and communicate complex ideas clearly to technical and non-technical audiences.
Ability to independently drive projects, take accountability for results, and focus on customer impact.

Responsibilities

Contribute to a high-impact research agenda in large language models and agentic systems, including data synthesis, model architectures, training, evaluation, and production deployment.
Design and execute post-training experiments using methods such as supervised fine-tuning, RLHF, RLAIF, DPO, and emerging alignment techniques to improve reasoning, planning, and tool use.
Build evaluation benchmarks and methodologies for assessing model performance on complex, real-world enterprise tasks.
Develop scalable data synthesis and curation pipelines, including LLM-as-judge frameworks, synthetic data generation, and adversarial dataset construction.
Translate research insights into improvements to WRITER’s enterprise-grade language models and help shape its model architecture and training roadmap in collaboration with research engineering and product teams.
Publish and present original research at leading venues and represent WRITER in the scientific community.
Collaborate with researchers and engineers and help establish high standards for scientific rigor, experimental design, and research culture.
Drive research projects end-to-end, collaborate across teams, challenge conventional approaches, and take accountability for results.

Benefits

Generous paid time off and company holidays.
Medical, dental, and vision coverage for employees and their families.
16 weeks of paid parental leave for all parents.
Fertility and family planning support.
Early-detection cancer testing through Galleri.
Flexible spending account and dependent-care FSA options.
Health savings account for eligible plans, with a company contribution.
Annual wellness stipend for expenses such as gym, massage or chiropractic care, and personal training.
Annual learning and development stipend.
Company-wide and team off-sites.
Company stock options and 401(k).
Competitive compensation.