SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Modulr is a fintech company building an industry-leading API platform for embedded payments. The company enables businesses from SMEs to enterprises across the UK and Europe to efficiently pay-in, collect, and disburse funds instantly through various payment schemes, accounts, and card products. Modulr is regulated as an Authorised Electronic Money Institution in the UK and Netherlands, with direct participation in major payment networks including Faster Payments, Bacs, CHAPS, Open Banking, SEPA, and SWIFT. The company is backed by PayPal, FIS, and growth investors including General Atlantic and Blenheim Chalcot, with over 400 employees across offices in London, Edinburgh, Amsterdam, and Mumbai.
The Data Scientist role focuses on building machine learning, deep learning, and generative AI solutions for high-impact financial use cases. Key responsibilities include developing ML/DL models for fraud detection, risk scoring, and personalization; building and deploying generative AI applications using large language models; designing and implementing agentic AI workflows; working with LangChain and LangGraph for orchestration; using tools like Claude and Cursor/CoPilot for AI-assisted development; implementing monitoring via LangSmith; building APIs for model deployment; collaborating across teams; and maintaining MLOps pipelines and CI/CD workflows.
The ideal candidate has 2–3 years of data science, ML, or DL experience with strong Python skills and familiarity with ML frameworks such as PyTorch, TensorFlow, and XGBoost. Hands-on experience with generative AI, LLMs, and prompt engineering is required, along with knowledge of LangChain, LangGraph, and agentic AI concepts. Experience with tools like Claude, Cursor/CoPilot, and LangSmith for debugging and monitoring is expected. Candidates should have knowledge of APIs, FastAPI/Flask, Docker, and MLOps tools like MLflow and DVC. Fintech domain experience is preferred.
