Junior Data Scientist

Requirements

• Bachelor’s or Master’s degree in Data Science, Computer Science, Statistics, Mathematics, or 1 year+ professional experience
• (Desirable) Demonstrated ability to design and validate features using transactional and behavioural data from financial services
• Some hands-on experience building ML models in Python
• (Desirable) Experience building and improving AI agents for specific tasks
• Working SQL skills
• (Desirable) Payment risk, fraud detection, or credit decisioning is a plus
• Solid understanding of supervised and unsupervised learning, particularly binary classification, cross-validation, and common evaluation metrics
• (Desirable) Any exposure to AWS is a bonus
• Comfortable working with tabular data
• Interest in working with product and engineering teams
• Curious, proactive, and comfortable working in ambiguity in a scaling start-up environment
• Detail-oriented and consistent, comfortable with careful, repetitive data work

What the job involves

• Clean and prepare raw transaction data for analysis and labelling, using natural language processing (NLP) techniques to parse and normalise messy text fields such as merchant names and descriptions
• Aggregate transaction data into customer- and merchant-level views, and surface behavioural insights such as spending patterns and recurring payments that inform features and labels
• Own the creation of ground-truth datasets from transaction data: define and apply labelling standards, run quality checks, and resolve ambiguous cases. Where possible, speed up labelling with heuristics, rules, and simpler models, and measure their accuracy against manually labelled samples
• Help build and test transaction-categorisation logic, for example surfacing merchant information, applying labels, and recording confidence scores
• Use Python for data preparation, feature engineering, and model evaluation, supporting the team's machine learning (ML) work
• Help validate model outputs, mostly for binary classification tasks, using cross-validation and confusion matrices, and reviewing precision, recall, and error patterns to flag where a model underperforms
• Apply a solid understanding of statistics and hypothesis testing to spot trends and support model development
• Work closely with cross-functional teams across product, engineering, and business stakeholders to understand requirements and deliver data-driven solutions
• Enthusiasm for collaborating across disciplines, including with academic researchers and third-party partners