Director Model Risk Management - AI/Gen AI

The Hartford•Hartford, CT

About The Position

The Hartford’s Model Risk Management function is seeking a director to join its talented and high-performing team. This role involves leading efforts to ensure the integrity, accuracy, and compliance of AI and Generative AI (GenAI) models used across the enterprise. The Director/Validator will independently review, challenge, and validate models to ensure they meet internal model risk management standards, regulatory expectations, and ethical AI principles. Additionally, the Director will drive the enhancement of the existing model validation framework for GenAI, including identifying and deploying model validation tools for increased efficiency. The Hartford utilizes advanced analytics, predictive, AI/ML, and Generative AI models, as well as traditional actuarial models, in various important and critical business functions. The Model Risk Management team manages model risk across The Hartford by validating these models, implementing consistent policies and standards, and maintaining appropriate model oversight. As part of this team, the role will primarily focus on validating AI and GenAI models across The Hartford and reporting results to key internal stakeholders. Additional responsibilities include educating on modeling best practices and spreading model risk awareness across the enterprise.

Requirements

Advanced degree (M.S. or Ph.D.) in a relevant field e.g., Artificial Intelligence, Machine Learning, Computational Science, Engineering, Statistics, Applied Mathematics, Actuarial Science, Quantitative Economics
10+ years of industry experience in machine learning or data science
1+ years focused on GenAI
Strong understanding of enterprise-wide governance and risk management frameworks
Strong programming experience across languages/technology platforms including Python, R, SAS/SQL
Solid understanding of GenAI concepts including prompt and context engineering, retrieval-augmented generation (RAG), agent workflow, LLM evaluation, familiarity with neural networks
Experience in GenAI tools such as Vertex AI/Google agent development kit, LangChain/LangGraph, RAG frameworks, HuggingFace, OpenAI APIs
Ability to act independently with proactive self-directed accountability and demonstrated experience and consistency in meeting deadlines while adapting to shifting priorities
Strong analytical, critical, and investigative thinking skills
Demonstrated commitment to lifelong learning with an ardent desire for continuous development to keep pace with evolving modeling techniques and AI technologies
Solution oriented creativity, innovative thinking, and challenging the status quo
Excellent communication and collaboration skills, with the ability to explain complex technical concepts to non-technical stakeholders across the enterprise

Nice To Haves

P&C, Group, Life, or related insurance product experience is a plus

Responsibilities

Direct and perform end-to-end model validations on AI and GenAI model use cases across The Hartford’s functional areas and lines of business
Ensure model calculations, machine learning algorithms, and GenAI methods are accurate and appropriate for intended use
Design and build challenger solutions and testing methods for tasks such as summarization, question answering, search, data synthesis, LLM-as-a-judge, Context Relevancy, Answer Relevancy, Groundedness etc.
Review and assess the quantitative and qualitative testing techniques to ensure model accuracy, robustness, and reliability
Assess key data inputs, assumptions, prompt engineering, context engineering for accuracy and appropriateness
Review model outputs for accuracy and appropriate downstream usage
Deliver effective challenge to key modeling elements such as inputs, calculations, outputs, conceptual soundness, monitoring & controls, documentation, etc.
Assess the appropriate use of model / use case controls, e.g., Guardrails, HITL/HOTL, their implementation and effectiveness across a variety of models and use cases
Identify findings and recommendations, including impact analysis, to mitigate model risk and compile clear and concise model validation reports
Perform governance accountabilities related to findings tracking, remediation testing, and validation
Drive end-to-end initiatives including the enhancement of the existing GenAI model validation framework
Assist in the continuous improvement of The Hartford’s Model Risk Management function by monitoring external environment, recommending process improvements, implementing emerging best practices, and evolving the enterprise’s model risk management Policy and Standards for Model Development and Use
Identifying and deploying model validation tools for increased efficiency, while ensuring the continued alignment with regulatory standards
Identify/develop qualitative assessments and quantitative performance metrics to test and monitor AI/ML and GenAI performance and reliability, including model drift detection, data currency, lineage, quality, integrity, and inform model validation practices (e.g., scope, frequency)
Pro-actively stay informed of advancements in AI/ML, GenAI modeling and associated emerging techniques/technologies, their application, risks, and risk mitigating strategies
Lead initiatives to understand and upskill for tools, such as VertexAI/Google agent development kit, LangChain/LangGraph, RAG frameworks, HuggingFace, OpenAI APIs, etc.
Strengthen enterprise partnerships with leadership and their teams across Data Science, Tech, PIDA, Actuarial and the Lines of Business
Deliver insights that enhance model development, performance, and reliability, ensuring a comprehensive approach to risk management and business strategy
Keep model risk practices aligned with the proliferation and sophistication of modeling by partnering on cross functional teams (e.g., Audit Readiness) to advance Standard Work Templates and best practices for proactive model risk management
Pro-actively stay informed of enterprise and Line of Business initiatives, deliverables, and reporting