Staff Applied ML Engineer - Financial Crime

About the role:

Wise moves billions across borders every year. Behind every transaction is a decision: is this safe? Our ML systems make that call - at scale, in real time, across every market we operate in.

Our Risk ML team is building the next generation of financial crime detection at Wise - investing in modern architectures like deep learning, graph neural networks, and foundation models to detect increasingly sophisticated fraud and money laundering patterns. We're looking for a Staff Applied ML Engineer to lead this evolution: defining the architecture strategy, shipping production neural models, and building the blueprint that scales across FinCrime domains.

This is a greenfield opportunity - you'll be setting the direction for how Wise applies modern ML to financial crime risk, with strong investment and engagement from senior leadership.

How we work:

Risk ML sits within Wise's FinCrime organisation, owning the full ML and AI foundation for financial crime detection. We're scaling into three dedicated pillars - Feature Platform, Learning Loop and Risk Modelling. You'll sit in Risk Modelling, working alongside data scientists, platform engineers, product and domain experts.

We operate with high autonomy and low hierarchy. You'll own problems end-to-end - from research and architecture decisions through to production deployment and impact measurement. We value engineers who shape direction, not just execute tickets.

What will you be working on?

• Designing and shipping ML and deep learning models for financial crime detection - sequence-based, graph-based, attention-based - serving real-time decisions at Wise's scale
• Defining the architecture strategy for how Wise applies modern ML to risk - which model families, which serving patterns, which training paradigms
• Building the reusable end-to-end pipeline pattern - from experimentation through training to production deployment - that future models follow
• Evaluating and prototyping foundation model and embedding approaches for transaction representation across FinCrime domains
• Partnering with Data Science on model evaluation, experimentation design and causal measurement in domains where clean A/B testing isn't always possible
• Mentoring engineers and data scientists on modern ML fundamentals, production best practices, and architectural decision-making

What do you need?

• Production experience shipping deep learning models at scale - systems serving real traffic under latency constraints
• Ability to make architecture-level decisions independently - model selection, training infrastructure, serving strategy - and explain the reasoning and tradeoffs
• Experience designing ML systems with hard latency and throughput requirements, including optimisation decisions (quantization, pre-computed embeddings, batching strategies)