Director, Data Science

Soteria Reinsurance Ltd.; Jersey City (NJ); On-site; USD 183,000 - 212,000; Full time

Job summary

Fidelity is seeking a leader to drive advanced AI solutions for financial services, focusing on LLM-powered applications and Agentic AI. You will design, train, fine-tune, and deploy transformer-based models within scalable ML infrastructure and LLMOps practices.

You will guide cross-functional teams, align with product and business stakeholders, and deliver production-ready AI features that accelerate growth while ensuring reliability and compliance.

Qualifications

• Bachelor’s degree in CS/Engineering or related field with 6 years in AI/ML leadership.
• Alternatively Master’s with 4 years in end-to-end AI/ML in financial services.
• Or PhD with 1 year in end-to-end AI/ML in financial services.

Responsibilities

• Oversees LLM post-training activities including fine-tuning, alignment, optimization, and evaluation.
• Architects and optimizes distributed ML training and inference systems for performance.
• Manages model hosting, serving, and deployment pipelines in cloud environments.
• Applies Agentic AI, prompt-engineering, and workflow orchestration for advanced AI apps.
• Develops synthetic data pipelines, evaluation methodologies, and custom loss functions.
• Guides cross-functional teams to deliver AI-enabled features and products.
• Establishes LLMOps and MLOps practices including monitoring and versioning.

Skills

Python

PyTorch

Transformers

CUDA

ML workflows

Agentic AI

LangGraph

ADK

Cloud ML Ops

MLOps

vLLM

PyTorch Profiler

EKS

Docker

Kubernetes

Education

Bachelor's degree in Computer Science/Engineering

Master's degree in Computer Science/Engineering

PhD in Computer Science/Engineering

Tools

Python

PyTorch

Transformers

CUDA

EKS

Docker

Kubernetes

vLLM

LangGraph

ADK

PyTorch Profiler

Job description

Job Description

Note: Fidelity will not provide immigration sponsorship for this position.

Position Description

Leads the development and delivery of advanced Artificial Intelligence (AI) solutions for financial services, with a particular focus on Large Language Model (LLM) powered applications. Provides strategic and technical leadership in designing, training, fine-tuning, and deploying transformer-based models and Agentic AI systems. Works within an innovation-driven incubator environment to guide end-to-end product development, from ideation and experimentation through production-ready implementation, using scalable Machine Learning (ML) infrastructure, distributed training systems, and modern Large Language Model Operations (LLMOps) practices.

Designs and implements scalable ML workflows using Python, PyTorch, and related frameworks to support production grade systems. Drives alignment with engineering, product, and business stakeholders to advance AI capabilities, accelerate experimentation, and deliver high-impact solutions that support new business growth across the financial services ecosystem.

Primary Responsibilities

• Oversees LLM post training activities -- fine tuning, alignment, optimization, and performance evaluation.
• Architects and optimizes distributed ML training and inference systems to enhance performance, memory efficiency, and latency.
• Manages model hosting, serving, and deployment pipelines using containerization technologies and model serving frameworks within cloud environments.
• Applies Agentic AI technologies, prompt-engineering methods, and workflow orchestration tools to develop advanced AI-driven applications.
• Develops synthetic data pipelines, evaluation methodologies, and custom loss functions to strengthen model robustness and accuracy.
• Guides cross-functional teams in delivering AI-enabled features and products, ensuring scalability, reliability, and compliance.
• Partners with product and business leaders to translate conceptual ideas into validated prototypes and production-ready solutions.
• Establishes and maintains LLMOps and Machine Learning Operations (MLOps) best practices, including monitoring, versioning, and continuous optimization of AI systems.

Education and Experience

Bachelor’s degree in Computer Science, Engineering, Information Technology, Information Systems, Physics, or a closely related field (or foreign education equivalent) and six (6) years of experience as a Director, Data Science (or closely related occupation) developing end-to-end AI and ML solutions, including LLM-based and Agentic AI systems, using Python, PyTorch, and cloud-based MLOps and LLMOps technologies within a financial services environment.

Or, alternatively, Master’s degree in Computer Science, Engineering, Information Technology, Information Systems, Physics, or a closely related field (or foreign education equivalent) and four (4) years of experience as a Director, Data Science (or closely related occupation) developing end-to-end AI and ML solutions, including LLM-based and Agentic AI systems, using Python, PyTorch, and cloud-based MLOps and LLMOps technologies within a financial services environment.

Or, alternatively, PhD in Computer Science, Engineering, Information Technology, Information Systems, Physics, or a closely related field (or foreign education equivalent) and one (1) year of experience as a Director, Data Science (or closely related occupation) developing end-to-end AI and ML solutions, including LLM-based and Agentic AI systems, using Python, PyTorch, and cloud-based MLOps and LLMOps technologies within a financial services environment.

Skills and Knowledge

• Candidate must also possess: Demonstrated Expertise ("DE") designing and orchestrating scalable ML workflows using Python, PyTorch, Transformers, Compute Unified Device Architecture (CUDA), ML algorithms, and workflow automation frameworks within a production grade AI development environment.
• DE architecting and optimizing distributed training and inference systems using parallel computed frameworks vLLM and leveraging performance-engineering, and profiling tools including PyTorch Profiler and Nvidia Nsight Systems within latency-sensitive AI environments.
• DE implementing model deployment, hosting, and serving using containerization technologies, model serving frameworks, and cloud infrastructure services (Amazon Elastic Kubernetes Service (EKS)) within a scalable enterprise production environment.
• DE advancing AI-driven application development using Agentic AI technologies (Agentic frameworks LangGraph and Agent Developer Kit (ADK)), prompt and context engineering methodologies, synthetic data pipelines, and evaluation frameworks within an Agile innovation and product development environment.

Salary: $183,000.00 to $212,000.00/year.

Fidelity’s Onsite Working Model

Fidelity is transitioning to a full-time onsite working model through a phased rollout across regions and roles. Currently, some roles and locations require 100% onsite presence, while others require less. Onsite expectations are likely to evolve as the rollout continues. This transition does not apply to fully remote roles.

Certifications

Category: Data Analytics and Insights

Regulatory Landscape

Please be advised that Fidelity’s business is governed by the provisions of the Securities Exchange Act of 1934, the Investment Advisers Act of 1940, the Investment Company Act of 1940, ERISA, numerous state laws governing securities, investment and retirement-related financial activities and the rules and regulations of numerous self-regulatory organizations, including FINRA, among others. Those laws and regulations may restrict Fidelity from hiring and/or associating with individuals with certain Criminal Histories.

At Fidelity, we are passionate about making our financial expertise broadly accessible and effective in helping people live the lives they want! We are a privately held company that places a high degree of value in creating and nurturing a work environment that attracts the best talent and reflects our commitment to our associates. We are proud of our diverse and inclusive workplace where we respect and value our associates for their unique perspectives and experiences.

Accommodation

Fidelity will reasonably accommodate applicants with disabilities who need adjustments to participate in the application or interview process. To initiate a request for an accommodation please contact the following: For roles based in the US: Contact the HR Leave of Absence/Accommodation Team by sending an email to [e-mail], or by calling 800-835-5099, prompt 2, option 2. For roles based in Ireland: Contact [e-mail]. For roles based in Germany: Contact [e-mail]. Fidelity Privacy Policy