Validateur de Modèles IA

Job Description

AI Model Validator | BNP Paribas

The Tone:
This is a full-time hybrid role at BNP Paribas, located in Montreal, Quebec, Canada. BNP Paribas, a leading international financial group and the first bank of the European Union, is dedicated to building a sustainable future. This role is crucial as part of the second line of defense for model risk management, ensuring the independent validation of artificial intelligence models across the group to comply with internal and regulatory standards. You will contribute to a dynamic, growing organization recognized as one of Montreal’s Top Employers 2025.

The TL;DR
• Role: Full Time
• Type: Full-time
• Location: Hybrid, Montreal, Quebec, Canada

• Team: Independent Review and Control (RISK IRC) team, a special unit within RISK, reporting directly to the Group CRO.
• Mission: Independently validate AI models used across BNP Paribas, ensuring their conceptual soundness, result validity, and adherence to internal and regulatory standards.
• Tech Stack: Python, R, GitHub, TensorFlow, PyTorch, NLTK, Gensim, GPT, Llama, Mistral, LangChain, LangGraph

What You’ll Actually Do
• Assessment: Perform independent quantitative evaluations of AI models utilized within BNP Paribas – CUSO IHC, adhering to internal standards and regulatory guidelines.
• Verification: Verify the conceptual soundness, validity of model results, implementation, and other relevant modeling aspects impacting model usage.
• Collaboration: Work closely with validation managers to develop appropriate validation plans that are proportional to the model’s risk level.
• Documentation: Prepare documentation for the process and audit materials for evaluation and decision-making committees.
• Presentation: Evaluate model performance and defend conclusions to management.

The Must-Haves
• Background: Graduate degree (Master’s/Ph.D.) in Finance, Economics, Statistics, Computer Science, or any other quantitative analysis-related field.
• Experience: Over 3 years of experience in AI model validation or development related to quantitative finance, such as derivatives, asset pricing modeling, or algorithmic trading.
• Skills: In-depth quantitative, statistical, and AI knowledge (predictive modeling, supervised/unsupervised learning, time series analysis, Natural Language Processing (NLP), and data science techniques), programming experience in statistical or mathematical languages like Python or R, practical experience with version control systems like GitHub, and experience with machine learning and NLP environments such as TensorFlow, PyTorch, NLTK, and Gensim. Professional English proficiency is required, as is the ability to write technical documentation and synthesize information through good written and oral communication.
• Bonus: Knowledge of regulatory requirements (e.g., SR26-2), familiarity with trading room products and strategies, real-life experience in developing/using machine learning projects, knowledge of cutting-edge Generative AI models (GPT, Llama, Mistral) and orchestration frameworks (LangChain and LangGraph), and an understanding of the technical, ethical, and regulatory aspects of AI and models.