Data Scientist

Work mode: Hybrid, 3 days WFO. Contract duration: 6 months. Location: London, UK. Start date of the project: 1st of September 2026.

Job Details

Role Title: Data Scientist (2 OPENINGS)

Years of experience: 3-5 years

Grade & Definition: Consultant – Substantial experience between 3-5 years and an in-depth knowledge of their specialist field.

Purpose of the Role

We are looking for talented and experienced data scientists with experience to join our programme. Solid knowledge and experience of AI and ML is essential. To work with the existing delivery team to deliver models, documentation and associated productionised services.

Responsibilities

• Design and develop AI / ML based solutions
• Work with other data scientists to build and deploy production-level solutions
• Troubleshoot and debug code
• Work with other teams to understand and solve business problems

Knowledge & Experience Requirements

Essential

• Python (pandas, NumPy, scikit-learn): For data wrangling, modelling, and feature engineering
• SQL: For querying structured data sources
• Model Development & Validation: Experience with classification, unsupervised learning (e.g. outlier detection), and ranking models
• Machine Learning Deployment: Familiarity with containerised deployment (e.g. Podman, SageMaker, DSW pipelines)
• Version Control (Git): To maintain reproducible and collaborative workflows
• Time-Series Analysis: To assess risk trends over financial years
• Exploratory Data Analysis (EDA): To spot early signals or risk clusters

Desirable / Advanced Skills

• Rank Aggregation/Ensemble Techniques: Understanding methods like Robust Rank Fusion (RRF)
• Model Explainability Tools: e.g. SHAP, LIME to support interpretability
• Experience with Model Monitoring & Drift Detection
• Experience in RegTech / FinCrime / Data-led Supervision Projects is a plus
• Experience developing solutions for record linkage and/or network analytics tasks
• Experience with graph query languages (e.g., Gremlin, Cypher), graph database platforms (e.g., Neptune, Neo4j), and/or graph visualisation platforms

#J-18808-Ljbffr…