Job Description

Data Scientist

PCD270007

Location: United Kingdom (Remote)
Type: 12 months Contract
Rate/Salary: 500 per day
IR35 Status: Outside IR35
Start Date: Jan 26
 

Role Summary

• Join a major UK public sector organisation to help expand data science and analytics capability across a high-impact portfolio of services.
• Deliver advanced analytical models and practical AI/ML solutions that support better decision-making and improved outcomes for the public.
• Work end-to-end across discovery, data acquisition, modelling, deployment, reporting and continuous improvement.
• Operate within a controlled and governed environment, producing clear, audit-friendly artefacts and defensible analytical outputs.
• Remote-first engagement with occasional site visits (as required for workshops, stakeholder sessions, or delivery milestones).

What You’ll Be Doing

• Lead the design and build of data science solutions from problem definition to production-ready delivery.
• Translate business and operational questions into measurable analytical objectives, hypotheses, and evaluation metrics.
• Develop and implement scalable data models and pipelines to support advanced analytics, forecasting, and reporting.
• Apply machine learning techniques to structured and unstructured data to uncover patterns, trends, anomalies, and predictive signals.
• Apply NLP and language model techniques to text-heavy datasets, including feature extraction, topic discovery, classification, summarisation, and semantic search use cases.
• Build and optimise workflows in partnership with Data Engineers, Analysts and Software Developers, ensuring solutions are reliable, repeatable, and maintainable.
• Integrate multiple data sources, improving data quality and consistency through validation rules, profiling, and data monitoring.
• Produce well-documented notebooks, model cards, assumptions, and technical documentation to support re-use, governance, and stakeholder confidence.
• Present findings and recommendations clearly, tailoring outputs for both technical audiences and senior stakeholders.
• Support longer-term capability uplift through reusable assets, documented standards, and pragmatic guidance on best practice.

Core Deliverables

• Data discovery outputs and problem statements with clear success criteria and measurable KPIs.
• Production-quality analytical datasets and feature sets, with repeatable pipelines and data quality checks.
• Machine learning models with robust evaluation, explainability considerations and clear operational assumptions.
• NLP / LLM prototypes and/or production workflows (where relevant), including embedding pipelines and retrieval strategies.
• Dashboards and reporting outputs to communicate insights (Power BI), aligned to stakeholder needs.
• Documentation suitable for a regulated environment, including technical specifications, runbooks, and handover materials.
• Ongoing model monitoring approach, including performance drift indicators and improvement backlog.

Key Responsibilities

• Own end-to-end delivery of analytical workstreams, including planning, prioritisation, estimation, and reporting of progress.
• Design and implement data pipelines and modelling workflows using Azure services and Python/R-based tooling.
• Build and evaluate statistical and machine learning models, selecting approaches appropriate for the problem and data maturity.
• Develop NLP pipelines for unstructured data, including clustering, embeddings, similarity matching, and classification tasks.
• Work with stakeholders to confirm requirements, validate findings, and ensure outputs are actionable and aligned to standards.
• Ensure consistent application of data governance principles including lineage, reproducibility, controlled access, and responsible use.
• Support CI/CD and MLOps practices for reliable deployments, automated testing, version control, and repeatable releases.

Technical Skills and Experience Required

• Strong hands-on data science experience delivering real outcomes (ideally within government, public sector, or regulated environments).
• Advanced proficiency in Python and/or R, including strong software engineering discipline (clean, secure, testable code).
• Strong foundation in statistics, experimentation, and machine learning, with the ability to justify model choices and assumptions.
• Experience working within the Azure data and analytics ecosystem, including:
  • Azure Machine Learning (AML) for training, model management, and deployment workflows
  • Azure Data Factory (ADF) for orchestration and data movement
  • Azure Data Lake / storage patterns for analytical datasets
  • Azure Synapse for querying and transformation (where applicable)
  • Awareness of Analysis Services and enterprise semantic models (beneficial)
• Strong data visualisation and insight storytelling skills using Power BI, including DAX and Power Query.
• Experience working with large, complex datasets and multiple upstream sources, including reconciliation, quality controls and consistency checks.

NLP, LLMs and Advanced AI (Important)

• Demonstrable experience using NLP techniques, such as:
  • Text preprocessing and feature extraction (tokenisation, n-grams, TF-IDF, etc.)
  • Clustering methods and topic discovery
  • Text embeddings and semantic similarity approaches
  • Classification and information extraction
• Practical experience with modern LLM patterns, ideally including:
  • Embedding pipelines and retrieval approaches (RAG-style patterns where appropriate)
  • Prompting strategies and evaluation approaches for quality and consistency
  • Orchestration frameworks such as LangChain (or equivalent)
  • Considerations for deployment, monitoring, and safe usage of LLMs in production settings

Desirable Experience

• Data pipeline automation, governance and operating-model maturity within large-scale environments.
• Familiarity with modern open-source AI frameworks and toolkits used in production data science (e.g., common ML/NLP libraries, model tracking, experiment management).
• Experience producing audit-ready analysis, including traceability of assumptions, data lineage, and reproducible experiments.
• Experience partnering with engineering teams to operationalise models (APIs, batch pipelines, event-driven processing).
• Experience working on transformation programmes where data quality and standards are evolving.