CCB Risk Program Associate

JPMorgan Chase & Co.•Wilmington, DE

About The Position

The Portfolio Risk Modeling team within CCB Risk Modeling group is responsible for end-to-end development of best in class forecasting model suite for Chase credit card portfolios to support stress testing, loss reserve, and business planning exercises. In this role, you, together with a team of highly-skilled quantitative professionals, will work on a large and cleanly structured codebase designed for large-scale distributed simulation and forecasting. Your expertise in machine learning, time series forecasting, causal inference and computer science will not only ensure us to deliver sophisticated models that are performant and in compliance with regulatory requirements and/or firm wide model risk policies but also enable our models and system run efficiently by writing effective and maintainable code. Additionally, you as a business professional will collaborate with business partners in loss forecasting, finance and technology, effectively communicate model results, analytical findings, and insights to them and senior leadership team to support business and or technical decisions.

Requirements

Master’s degree in Computer Science, Mathematics, Statistics, Econometrics, Physics, Engineering, or related quantitative fields.
2 years of experience with data analysis in Python.
Proven track record designing, building, and deploying high-quality machine learning models in production environments.
In-depth knowledge of advanced ML algorithms: logistic regressions, linear regressions, XGBoost, Deep Neural Networks (CNN/RNN), clustering, and recommendation systems.
Experience interpreting complex models (XGBoost, GBM, deep learning).
Familiarity with large language models, including fine-tuning and deployment for NLP tasks.
Minimum one year of hands-on experience with Python, TensorFlow, Spark, or Scala, and big data technologies (Hadoop, Teradata, AWS Cloud, Hive).

Nice To Haves

PhD in a quantitative field with publications in top journals, preferably in machine learning.
Strong expertise and research track record in Explainable AI (XAI) and LLMs.
Expertise in data wrangling and model building on distributed Spark environments with stability, scalability, and efficiency. GPU experiences desired.
Hands-on experience with LLM techniques: prompt engineering, fine-tuning, model distillation, and optimization (DPO, PPO).
Experience building agentic AI systems: tool-calling agents with retrieval, reasoning, secure execution (function calling, orchestration, policy enforcement) following MCP protocol, including safety and compliance guardrails.
Experience building RAG pipelines: domain knowledge curation, data-quality validation, and feedback loops for knowledge freshness.
Proven production implementation track record with strong ownership and execution.

Responsibilities

Design and develop machine learning models, time series forecasting models to support loss and revenue forecasting under regulatory and business framework.
Research, develop, document, implement, maintain, and support tools and frameworks that enhance AI/ML model explainability and fairness, ensuring transparency and ethical use of models.
Utilize state-of-the-art machine learning methodologies and construct sophisticated models, including deep learning architectures, on big data platforms to solve complex business challenges.
Design and implement tool-calling agents combining retrieval, structured reasoning, and secure action execution with robust guardrails for safety and compliance.
Curate domain knowledge, build data-quality validation frameworks, and establish feedback loops to maintain knowledge freshness.
Collaborate with diverse teams, including Loss Forecasting, Finance, Technology, Governance and Review, throughout the entire modeling lifecycle, from development and review to deployment and operational use.