About the Company
Thales UK is committed to delivering high-impact data science capabilities that enable data-driven decisions, improve the quality of our offers, win new business, and enhance customer satisfaction. Our culture fosters collaboration and innovation within a growing software and AI team in CortAIx Factory.
About the Role
The Data Scientist will collaborate with product owners, domain experts, data engineers, and software engineers to turn business problems into robust, secure, and scalable analytical solutions. This role will contribute reusable datasets, features, templates, and analytics components to Thales UK’s internal catalogue, accelerating the adoption and continuous evolution of data science across programmes.
Responsibilities
- Translate business questions into analytical problem statements, hypotheses, and measurable success criteria (KPIs, guardrails, value metrics).
- Lead data discovery and exploratory data analysis; profile data quality, completeness, bias, lineage, and document data dictionaries/metadata.
- Apply rigorous statistical methods: experimental design (A/B and multivariate tests), hypothesis testing, power analysis, causal inference and quasi-experimental techniques.
- Engineer robust features and curated datasets; contribute to feature stores and reproducible data pipelines with clear versioning and provenance.
- Quantify uncertainty and risk (confidence intervals, calibration, Monte Carlo simulation, scenario analysis) to support decisions in safety- and mission-critical contexts.
- Ensure Responsible Data and AI practices: explainability (e.g., SHAP/LIME), fairness/bias assessment, privacy-preserving methods, and alignment with MOD/regulatory and internal governance requirements.
- Build reproducible analysis artefacts (notebooks, scripts) and user-facing insights (dashboards, reports, briefings); automate recurring analyses and collaborate with engineers to productionise models and analytics.
- Define operational metrics and monitoring for model/data performance, data quality, and drift; support post-deployment reviews and continuous improvement.
- Produce clear technical documentation (analysis plans, model cards, experiment reports) to a high standard.
- Create reusable analytics components, templates, datasets, and reference implementations; contribute to the internal catalogue of capabilities.
- Support bids, PoCs, demos, and stakeholder workshops; communicate complex concepts and insights to non-technical audiences with compelling narratives and visuals.
- Work with data engineers and architects on data acquisition strategies, labelling/annotation approaches, integration of third-party data, and data quality management.
- Participate in agile threat modelling and vulnerability management for analytics solutions; adopt best practices for secure handling of sensitive data.
- Horizon scan for data science and analytics trends; run trials and share best practices to accelerate responsible adoption across programmes.
Qualifications
- 5+ years’ experience delivering data science solutions in complex, safety- or mission-critical domains (e.g., defence, aviation, rail, medical, or similar).
- Proven track record taking analytics from discovery and experiment design through modelling and validation to operational adoption (dashboards, decision support, or model deployment) with measurable outcomes.
- Significant hands-on experience in at least one area: time series forecasting, anomaly detection, optimisation/operations research, or NLP analytics.
- Demonstrated experience defining KPIs/measurement frameworks and conducting A/B or quasi-experimental studies, or causal inference for impact assessment.
- High-quality technical documentation and stakeholder communication.
- Collaboration within cross-functional engineering and product teams.
Required Skills
- Strong Python and/or R programming skills; proficiency with modern software engineering practices for data science (testing, code quality, CI).
- Solid grounding in statistics and probability: experimental design, inference, hypothesis testing, sampling, power analysis, regression/GLMs, Bayesian methods.
- Experience with core data science libraries: pandas, NumPy, scikit-learn, statsmodels; familiarity with PyTorch/TensorFlow and Hugging Face for NLP/LLM-assisted analytics is a plus.
- Data wrangling and querying (SQL); familiarity with Spark/Databricks or similar distributed data tools is desirable.
- Visualisation and storytelling: Matplotlib/Seaborn/Plotly; experience with BI tools (Power BI/Tableau) is desirable.
- Experiment tracking and reproducibility: MLflow, Weights & Biases, DVC; environment management (conda/poetry) and notebook best practices.
- Awareness of MLOps and data engineering: feature stores, workflow orchestration (Airflow/Prefect), API-based scoring (FastAPI), version control (Git).
- Responsible data and security awareness: data ethics, PII handling, GDPR/UK DPA, privacy-preserving methods (e.g., differential privacy, k-anonymity), explainability, and adversarial robustness basics.
- Knowledge of cloud data/ML services (AWS/Azure/GCP) and containers (Docker) is desirable; GPU acceleration awareness is a plus.
- Optional domain toolkits (a plus): time series (Prophet, darts), geospatial (GeoPandas), simulation/Monte Carlo, causal inference (DoWhy, EconML).
Preferred Skills
- Governance of analytics architecture/datasets and data catalogues throughout the project lifecycle.
- Experience with large-scale data initiatives, labelling strategies, and data quality management.
- Familiarity with MLOps practices and cloud platforms for analytics/ML (e.g., Databricks, Azure ML, SageMaker).
- Contributions to open-source projects, publications, or patents.
- Experience working with sensitive data and in regulated/MOD environments.
Pay range and compensation package
Competitive salary based on experience and qualifications.
Equal Opportunity Statement
Thales UK is committed to diversity and inclusivity in the workplace. We encourage applications from all qualified individuals regardless of race, gender, disability, or any other characteristic protected by law.
