Senior Applied AI Research Scientist

About the role

Sardine is building the world’s most advanced fraud‑detection platform and needs an applied AI researcher to push foundation‑model research forward. You will design, train, and deploy state‑of‑the‑art deep‑learning models that operate on massive behavioral and transactional data streams.

Key responsibilities

• Identify research opportunities, design rigorous experiments, and drive the roadmap for foundation‑model development.
• Own evaluation pipelines, including offline benchmarks, time‑aware holdouts, calibration, drift monitoring, and head‑to‑head comparisons with classical baselines.
• Take models from data preparation and tokenization through pre‑training, fine‑tuning, distillation, quantization, and real‑time inference deployment.
• Partner with engineering on training infrastructure, GPU efficiency, feature stores, and production‑scale serving.
• Collaborate with client‑facing, legal, compliance, and risk teams to ensure model explainability, documentation, and regulatory governance.

Required profile

• 4+ years of applied machine‑learning, quantitative modeling, or ML‑engineering experience, including at least one foundation model that you have pre‑trained or heavily adapted for live traffic.
• Hands‑on experience with self‑supervised pre‑training, fine‑tuning, and production model serving, versioning, monitoring, and rollback.

Required skills

• Strong Python programming.
• Advanced SQL for large‑scale data preparation.
• Experience with GPU‑accelerated training and efficient inference pipelines.
• Knowledge of foundation‑model architectures, deep learning, and sequence modeling.

What we offer

• Generous cash and equity compensation with early‑exercise options.
• Remote‑first culture – work from anywhere in the US or Canada.
• Flexible paid time off, health, dental, vision benefits, 401k/RRSP matching.
• Home‑office stipend, MacBook Pro, meal and wellness allowances, learning budget.