MLOps Developer, Risk

Your career is an investment that grows over time!

Wealthsimple is on a mission to help everyone achieve financial freedom by reimagining what it means to manage your money. Using smart technology, we take financial services that are often confusing, opaque and expensive and make them transparent and low-cost for everyone. We’re the largest fintech company in Canada, with over 3+ million users who trust us with more than $100 billion in assets.

Our teams ship often and make an impact with groundbreaking ideas. We're looking for talented people who keep it simple and value collaboration and humility as we continue to create inclusive and high-performing teams where people can be inspired to do their best work.

The quantitative risk team is responsible for the data and models that protect Wealthsimple and its customers from fraud, credit and market events. We are looking for a founding MLOps developer to lead the development of our AI/ML pipelines. This function plays a critical role in accelerating our adoption of AI, automating our training and tuning pipelines, and hardening our ML infrastructure against threats.

This individual will leverage a deep understanding of ML operations, engineering, and AI and the effective application of these tools in finance. They will apply this knowledge to enhance our end-to-end modeling infrastructure and help us deploy and maintain advanced models. This role offers the unique opportunity to work on high-impact projects that directly contribute to the security and efficiency of our platform.

• Automated Model Recalibration Pipeline: Implement and deploy an automated pipeline for model recalibration and re-fitting across key models, starting with those for fraud detection. Your work will ensure our models maintain optimal performance through systematic refreshing as market conditions evolve and threat actors' tactics change.
• Wealthsimple AI Foundation Model: Contribute to the continued development and productionization of our foundational AI model. This project requires both data science expertise and engineering rigor to transform how we understand our clients. You will apply advanced machine learning techniques while building scalable training pipelines and optimizing model performance. Your work will involve designing systems that effectively process diverse client signals (trades, transfers, clicks, purchases) into a unified high-dimensional embedding, then engineering these solutions from development to production.
  
  This versatile embedding will serve as the core intelligence layer for numerous downstream applications, including:
  • Fraud Detection: classify signals associated with known fraud types.
  • Client Segmentation: group clients based on their unique behaviours.
  • Outlier Detection: identify abnormal behaviours and patterns beyond traditional fraud, enhancing our ability to maintain a secure and trustworthy platform.
  • Model Explainability: Drive model explainability (xAI) initiatives to ensure ML models meet regulatory requirements and can be successfully deployed to production environments. Develop and implement interpretability techniques that make complex models transparent to stakeholders, satisfying compliance standards while enabling faster model approval and deployment cycles.
  • MLOps Security & Infrastructure Expertise: Lead the implementation of best practices in our machine learning operations to create a secure, scalable infrastructure. You will architect robust ML systems that protect against cybersecurity vulnerabilities including data poisoning, model evasion, and adversarial attacks, while ensuring our decision infrastructure remains performant and maintainable at scale.

• 5+ years of experience as a data scientist and/or machine learning engineer, with a focus on advanced models in finance.
• Expertise in Python, with a strong familiarity with key data science & machine learning libraries and tools. (PyTorch/Tensor Flow/Light
  
  GBM/Scikit-Learn).
• Ownership of the design and development of an ML platform: demonstrated experience in productionalizing ML models, with a track…