Senior Data Scientist

About the role

In 2018, Bitvavo launched the first version of its digital assets trading platform, to bridge the gap between traditional currencies and digital assets. By offering transparent fees, a wide range of assets and an easy to use platform, Bitvavo is making the currency of the future accessible for everyone.

In two years' time, Bitvavo has established itself as the market leader in the Netherlands and one of the largest in Europe, with over one million users exchanging tens of billions of digital assets a year.

ABOUT THE OPPORTUNITY

We are an exchange. The speed of good decisions matter in everything that we do. Our data domain is committed to making sure the decisions this business makes on spend, pricing, product, risk and financial crime rest on evidence rather than intuition. We're seeking exceptional talent to join us in making this a reality.

The Problems You'll Get to Solve

- Growth & causal inference. Measure true incrementality, own our marketing mix model, and turn that into spend allocation under a budget.

- Product & personalization. Rank and surface what is relevant to each customer's behaviour and holdings, from asset discovery to order form defaults, without drowning users in noise

- Markets, pricing & execution. Model fee and spread elasticity to find the revenue-versus-volume trade-off, and improve execution quality: slippage, order routing, liquidity provider selection.

- Risk & financial integrity. Model risk exposure across a volatile asset book, forecast liquidation risk, and detect reconciliation breaks across internal ledgers, on-chain data and partner reporting.

- Financial crime & compliance. Detect account takeover, payment fraud and mule networks in real time, combining business rules and models that work together while staying explainable and auditable.

What You'll Be Doing

- Own problems end to end: frame the question with the stakeholder, build the model or experiment, ship it, and follow it into production.

- Work backwards from the decision. Start from what someone needs to decide and what it costs to get it wrong, not from the data that happens to be available.

- Stay accountable for your models once they are live: monitoring, drift, retraining, and the day they start behaving differently.

- Tell senior stakeholders things they do not want to hear, including that a campaign, a feature or a rule did not work.

- Support your teammates in growing every day by setting the standard for how we do causal inference, reviewing their work, and sharing knowledge.

What We Look For

- Strong applied experience turning a messy business question into a model or a measurement: causal inference where that fits, predictive and ranking models where it does not. You know when a result is not real.

- Fluency in Python and SQL, and comfort working independently rather than waiting for data to be handed to you. We use Snowflake, dbt, Airflow and Looker, on top of technologies like MySQL, Druid, Kafka and Flink.