An early-stage lending company is seeking an experienced Data Scientist for a contract position focused on building a transaction categorisation system.

This is an opportunity to work at the intersection of open banking, NLP, and credit risk modelling, helping to standardise data ingestion from multiple banking data providers.

You’ll be involved in designing and training models to interpret and categorise bank transactions - transforming raw data into a unified, structured format.

Key Responsibilities

Design and implement ML/NLP-based models to categorise raw bank transaction data.
Standardise outputs across multiple open banking data providers into a consistent schema.
Evaluate discrepancies in categorisation logic from existing providers and design model features to mitigate them.
Iterate and improve models based on performance, accuracy, and real-world cases.
Ensure models are explainable and auditable, suitable for regulated environments.

Skills & Experience

Experience in machine learning, natural language processing, or data science, ideally in fintech, banking, or financial services.
Solid understanding of open banking, bank transaction data, or personal finance categorisation.
Hands-on experience with Python (and relevant libraries such as scikit-learn, pandas, spaCy, or similar) and SQL.
Experience working with raw, messy transactional data and transforming it into structured outputs.
Strong background in model training, and validation.
Familiarity with version control, pipelines, and productionising ML models is a plus.
Previous experience supporting or integrating with credit risk or affordability frameworks is highly desirable.

Contract Details

Location: UK, Remote

Engagement: Contract

Day Rate: Open to discussion based on experience

Start Date: ASAP