Remote job
Applied Scientist, Foundation Models
Job details
About this role
Role overview
Help establish a foundation-model capability for a consumer financial services platform. You’ll build or adapt models for domain-specific needs, bring them into production with partner teams, and improve search and retrieval while accounting for reliability and bias risks in a regulated setting.
Responsibilities
- Design, train, and evaluate foundation models using proprietary data or adaptations of existing open or commercial models. - Fine-tune and deploy models into production applications in partnership with product, engineering, and analytics teams. - Identify and reduce risks such as biased behavior and unreliable model outputs. - Design and optimize search and retrieval systems to improve relevance and help users find information. - Scope ambiguous problems, run focused experiments, and connect technical work to measurable business outcomes.
Requirements
- At least five years of experience building and productionizing machine-learning or AI models with measurable impact. - Hands-on depth in either large-scale foundation-model pretraining, including multi-GPU training, or domain-specific LLM fine-tuning; strong understanding of transformer architectures. - Solid grounding in traditional machine learning, with experience in areas such as credit, fraud, ranking, recommendations, or personalization. - Degree in computer science, data science, applied mathematics, or a related field. - Ability to work independently and collaborate across product, engineering, and analytics from early exploration through deployment.
Benefits and work setup
- U.S.-based remote-friendly work and flexible time off. - Base salary, stock options, health benefits from the first day, and a 401(k) plan with company match.