Data Scientist

On behalf of NHS Counter Fraud Authority, we are looking for a Data Scientist (Inside IR35), Band 8a for a 6 Month contract based REMOTELY with occasional meetings in London, Newcastle and Coventry.

NPPV2 Clearance is an essential requirement for this role, (as a minimum you must be eligible and willing to undergo these checks).

The NHS Counter Fraud Authority (NHSCFA) is the National body responsible all for matters relating to the prevention, detection and investigation of economic crime across the NHS. Aligned to the DH Health Group Counter Fraud strategy, the NHSCFA acts as the principal lead for the NHS and wider health group in counter fraud intelligence work.

Main responsibilities:

*Create, deploy, and refine cutting-edge AI/Machine Learning algorithms and suitable methodologies to detect and anticipate evolving fraud patterns and trends.

*Deliver expert knowledge and practical experience working with Alteryx, R, Python, and SQL, and an understanding of the concepts behind cloud computing technologies and efficient processing across the entire portfolio of work.

*Harness the benefits of a cloud-based platform to develop analysis, statistical and machine learning models, host interactive visualisations, and collaborate on reproducible coding with Git.

*Design innovative solutions to solve business problems in the identification of fraud and wider NHS to protect NHS money and resources from irregular activity.

*Build highly accurate practical Machine Learning models, by developing supervised, unsupervised, and semi supervised models etc, creating end-to-end data pipelines and deploying outputs within the NHSCFA environment ready for action.

*Prepare data for model development and selection using techniques such as, sampling, feature engineering and normalisation etc.

*Leverage advanced AI and machine learning techniques, including deep learning, to improve fraud detection and prevention models while adhering to privacy regulations.

*Provide a deep understanding of the theoretical foundations behind classical and recent machine learning models and algorithms, such as generalised linear models, random forests, SVM, ensemble methods, and deep neural networks etc including assessment and justification of approaches and metrics and how each is used in a practical environment to detect anomalies/fraud within the NHS including providing verbal and written explanation of the results and key metrics.

*Minimise false positives while ensuring predication or classification accuracy is paramount.

*Deploy the models in the operational environment and maintain/troubleshoot any production issues that arise.

*Participate in the full development cycle: design, develop, QA, experiment, analyse, and deploy models into the NHSCFA environment ensuring each phase is appropriate, documented, and transparent, aligned to government standards.

*Collect, process, analyse and disseminate data to support statistical analysis and risk-based decisions made using the various deployed models and methods which generate insight.

*Write high quality production ready and reusable code and packages for the models deployed in the production environment.

*Develop key metrics to track the detection and performance of fraud/anomalies and verification strategies building transparency from model selection through to evaluation and dissemination.

*Apply statistical, and operational research techniques to effectively model and analyse a variety of data sources, and make recommendations based on findings to support preventative, operational and strategic business decisions.

*Present technically complex findings to multi-disciplinary groups of varying sizes in a clear, concise, and engaging way. Responding tactfully to challenge of assumptions and results when questioned.

*Explain model concepts, output/results, anomalies and metrics to those involved in the portfolio of work, stakeholders and assurance validators justifying approach.

Knowledge and Experience

Essential

*Practical application experience in managing data science projects, from project design, method selection, control, optimisation, and implementation within the workplace. Including guidance, documentation, and leadership.

*Proven experience of line managing a team of analyst as part of a data science project.

*Knowledge and practical experience in designing algorithms including selection, using statistical and problem centric methods to design actionable outcome across a variety of data types.

*Proven practical experience and expertise in AI/machine learning methods using data including the implementation of statistical modelling and methods into a live production environment

*Experience and strong proficiency in programming languages for data science, e.g., SQL, R and Python alongside the ability to use tools and packages such as Alteryx, Jupyter notebook, R Markdown, TensorFlow, Keras, Pytorch etc.

*Practical expertise in producing reproducible code and pipelines including documentation, governance and assurance frameworks, automation and code review using tools such as Git.

*Skilled in data visualisation, with expertise in employing best practices and a proven track record of effectively communicating statistical findings to stakeholders.

*Strong analytical and problem-solving skills, with the ability to analyse large and complex datasets, extract meaningful insight and actionable outcome to help inform business outcome.

*Present technically complex findings to multi-disciplinary groups of varying sizes in a clear, concise, and engaging way. Responding tactfully to challenge of assumptions and results when questioned.

*Explain model concepts, output/results, anomalies and metrics to those involved in the portfolio of work, stakeholders and assurance validators justifying approach.

*Practical experience in model development and production and the alignment with model governance, business, best practice, and regulatory standards.

*Practical and proven expertise of machine learning algorithms, including supervised, unsupervised, and semi supervised techniques used to build and deploy models.

*Significant expertise of techniques used to prepare data for model selection, such as sampling, normalisation, imputation including theoretical knowledge. .

*Proven experience in operating within a big data environment, encompassing the management of substantial data volumes sourced from diverse origins, including data handling, pre-processing, scalability, storage optimisation, and the application of efficient processing methodologies.

Desirable

*Expertise and proven experience of fraud/anomaly detection

*Extensive understanding of the NHS data landscape.

*Accredited Counter Fraud Specialist or member of Government Counter fraud Profession

*Knowledge of generative AI models involving NLP across both open and closed large language models.

*Domain knowledge of key NHSCFA thematic areas and corresponding datasets across the NHS, wider public sector and beyond.

*Practical expertise of machine learning algorithms used to detect fraudulent activity.

Please be aware that this role can only be worked within the UK and not Overseas.