Analyst/Associate/Assistant Vice President - Model Validation - Artificial Intelligence Applications & Large Language Models

Description:

Levels - Analyst/Associate/AVP

Qualification, Experience & Skills: Essential

- Bachelors degree in Computer Science, Data Science, Engineering, Mathematics, or related field

- 1-7 years experience in data science or quantitative analysis or text analytics

- Knowledge of Gen AI, LLMs concepts, RAGs, agentic AI frameworks, and their applications

- Experience in scientific programming & data visualization in Python and its libraries

Business Unit Overview:

Model Risk Management is a group within Risk Management responsible for:

- Executing and maintaining an effective Model Risk management framework.

- Producing a consolidated view of Model Risk for comparison with the Model Risk Appetite.

- Independently validating the integrity and comprehensiveness of the firm's Models.

Due to the expanding scope of Model Risk Management, we are seeking a Model Validation Associate specializing in Artificial Intelligence Applications and Large Language Models (LLM). The successful candidate will have a strong technical background and will be responsible for the independent validation of AI Models and technologies across various business functions. Use cases include chatbots, sentiment analysis, translation tools, voice recognition, computer vision, data extraction, and other advanced applications and LLMs.

The role will specifically cover the following areas:

Independent Validation of AI Models, including:

- Assessment of the conceptual soundness of AI Models and/or Use Cases, including the integrity and suitability of AI parameters and prompts.

- Model Risk Analysis identifying, analyzing, and quantifying AI Model Risks such as bias, fairness, and hallucination, involving independent testing against business objectives.

- Developing Model interpretability and explainability frameworks (e.g., SHAP, LIME, attention mechanisms).

- Conducting Robustness testing against distribution shifts and edge cases.

- Development and Execution of Model Performance Monitoring to ensure AI Models perform as intended without drift, hallucination, or biased outputs.

- Design and implementation of Model Risk Control processes for AI Use Cases.

- Preparation of comprehensive model review documentation.

- Supporting risk teams in understanding and managing AI-related Risks.