Deuna Data Scientist role at Deuna to lead machine learning and experimentation for payment optimization. Responsible for designing, training, and validating models to improve conversion, reduce declines, and reduce fraud in real-time payment flows.
Responsibilities
We are looking for a hands-on data science leader to own machine learning and experimentation for DEUNA’s payments platform. You will be building the foundations of the data science practice at Deuna, working independently to design, train, and validate models that increase conversion, reduce declines, and reduce fraud in real-time payment flows across merchants and payment providers. You will be a part of the technology organization and will collaborate with project leads who own AI and modeling solutions for our customers, alongside existing data engineering and BI teams who already manage our data foundations and inference services. As DEUNA’s AI efforts grow, so will the team, and this role is a technical leadership role, with potential for people leadership if the skills align.
Own the end-to-end modeling lifecycle for DEUNA’s payment optimization use cases: analyze merchant and transaction data, define modeling strategy, generate features, train and tune models, and release them through our existing inference infrastructure.
Build the experimentation framework we need to measure model performance while collaborating with system and infra engineers, from early signals through month-over-month results, and tie that performance to commitments on our merchant contracts and product pipeline.
Partner with data teams to define and track building of data processing pipelines that surface insights that shape what DEUNA delivers. You will use experiments and rapid model iteration to tailor approaches to merchant data and goals, and work with the existing models while building new ones.
Qualification
You think like a research scientistYou have designed and run experimentsYou are self-directed
Required
You have 5+ years of experience applying machine learning to real production problems, ideally in payments, fintech, fraud, or other high-volume transactional environments.
You think like a research scientist: you start from a problem, decide what data you need, choose the right modeling approach, and define how you will prove it works.
You have hands-on depth in feature generation, model selection, training, and hyperparameter optimization, and you know when a simple model beats a complex one.
You have designed and run experiments, including A/B testing and experimentation platforms, and can explain how you validated that a model delivered impact.
You are proficient in Python and SQL and comfortable querying large datasets in tools like Snowflake. Familiarity with tools such as TensorFlow and Airflow is expected, and you can explain how you have managed and rolled out models.
You are self-directed. You push for the information you need, and you do not need close supervision to keep work moving.
You communicate clearly with technical and non-technical audiences, including setting realistic expectations with leadership about what models can and cannot achieve.
You have experience in guiding technical infrastructure teams to implement features that enable efficient data science and ML training pipelines and systems
