Applied AI Research Scientist

Job description

**This position is listed on behalf of a partner company, who manages all applications and next steps. Our partner is looking for an Applied AI Research Scientist based in Canada.** As an Applied AI Research Scientist, you’ll apply deep learning and foundation model expertise to some of the most challenging problems in fraud detection and financial risk. You’ll work with rich, large-scale behavioral and sequential datasets spanning device intelligence, biometrics, telemetry, payments, and consortium signals. The role combines cutting-edge applied research with hands-on production engineering, taking models from experimentation through real-time deployment. You’ll help define and execute the roadmap for next-generation fraud foundation models while establishing rigorous standards for model evaluation and performance. Collaboration will span data science, engineering, product, compliance, legal, and customer-facing teams. You’ll also work directly with financial institutions and fintech organizations to translate model capabilities into practical risk decisions. This is a remote-first opportunity for an independent, highly motivated scientist who wants their research to create measurable real-world impact. ### Accountabilities - Identify high-value opportunities for foundation model research and development, scope initiatives, design rigorous experiments, and drive execution against the research roadmap. - Develop next-generation fraud detection solutions by applying foundation models, deep learning, and representation learning to large-scale non-text sequential data. - Establish and maintain a high evaluation standard through offline benchmarks, time- and entity-aware holdouts, calibration analysis, drift monitoring, degradation tracking, and comparisons against strong classical baselines. - Take models through the complete machine learning lifecycle, including data preparation, tokenization, pretraining, fine-tuning, distillation, quantization, deployment, and production optimization. - Partner with engineering teams on training infrastructure, GPU utilization and efficiency, feature and embedding stores, model serving, and scalable real-time inference. - Ensure models meet demanding production requirements, including tight latency constraints, reliable serving, version control, monitoring, and rollback capabilities. - Collaborate with client-facing teams and customers to communicate model capabilities, limitations, and performance in ways that enable actionable risk-management decisions. - Work with legal, compliance, and customer model-risk teams to develop appropriate explainability, documentation, governance, and validation practices for regulated financial environments. - Independently manage ambiguous applied research projects while maintaining clear communication and alignment with internal and external stakeholders. - Contribute to the advancement and adoption of practical, state-of-the-art AI approaches within fraud detection and financial risk. ## Requirements - 4+ years of experience in applied machine learning, quantitative modeling, ML engineering, or a closely related field. - Hands-on experience pretraining or substantially adapting at least one foundation model and deploying it in a production environment with real-world traffic. - Strong practical experience with self-supervised pretraining, fine-tuning, and model adaptation techniques. - Production experience with model serving, versioning, monitoring, and rollback processes. - Strong Python and SQL skills, with demonstrated ability to prepare, process, and analyze very large datasets. - Ability to independently scope and execute ambiguous research and development projects from experimentation through production. - Excellent communication and collaboration skills, with the ability to work effectively across data science, engineering, product, marketing, compliance, legal, and external partner teams. - Strong research mindset combined with a practical, outcome-oriented approach to building production-ready machine learning systems. - Experience in fraud, AML, payments, credit, or adversarial machine learning is a strong asset. - Experience building and evaluating LLM-based agents in production is a plus. - Publications, released models, or open-source contributions involving representation learning or sequence modeling are advantageous. - Experience with model risk management, regulatory documentation, and governance in financial services is a plus. ## Benefits - Generous compensation package combining cash and equity. - Early exercise opportunities for all stock options, including pre-vested options. - Remote-first culture with the flexibility to work from anywhere. - Flexible paid time off and a year-end company break. - Health, dental, and vision insurance for employees and eligible dependents in Canada. - 4% RRSP matching for eligible Canadian employees. - MacBook Pro provided for work. - One-time home office setup stipend for equipment such as a desk, chair, and monitor. - Monthly meal stipend. - Monthly social meetup stipend. - Annual health and wellness stipend. - Annual learning and professional development stipend. - Opportunity to work alongside experienced professionals on complex, high-impact AI and financial risk challenges. - Globally distributed, collaborative environment with a strong emphasis on autonomy, ownership, and results rather than hours worked.