Data Scientist

Description ICF supports governments and organisations on challenges in areas such as energy, the environment, aviation, health and public policy. Our data science team works with subject matter experts to turn evidence and data into practical solutions for clients in the UK and EU. What you will do Speak with subject matter experts and clients to define the problem, success measures, data needs and an appropriate analytical approach. Prepare and analyse data, develop and evaluate models or prototypes using statistical methods, forecasting, machine learning or language models where suitable. Present findings, assumptions and limitations to subject matter experts, clients and nontechnical audiences, invite challenge and adapt the work to their feedback. Collaborate with colleagues across different technical disciplines to document work, make results reproducible and develop promising prototypes into reliable solutions. Contribute technical ideas, estimates and evidence to proposals, and manage your work across several concurrent projects. Share methods and reusable approaches with colleagues across the data science team. What you will bring Hands‑on experience in Python for analysis and modelling, including common data and machine learning libraries, ability to use SQL to extract, join and check data. Sound grounding in statistics and machine learning, including regression or classification, forecasting, validation, baseline comparisons and appropriate performance measures. Practical familiarity with using large language models in analysis or prototyping, including prompting, checking outputs against evidence and recognising privacy, accuracy and reliability risks. Experience preparing imperfect data, building reproducible analyses and writing maintainable code with version control and testing. Ability to take an ambiguous question from discovery to a working demonstration, explain assumptions and uncertainty, and judge whether a solution is worth developing for operational use. Confidence working independently and engaging stakeholders across teams, clear written and spoken English for technical and nontechnical audiences. A relevant degree or equivalent practical experience in a quantitative field. Useful additional experience Experience working in public policy, government, research or evaluation Experience in consulting, public policy research or client‑facing delivery, familiarity with cloud platforms, APIs, data pipelines or deployment. Experience building AI agents or applications that connect language models to data and tools, such as retrieval‑based systems. We can train someone with strong core skills. What you can expect You will work alongside colleagues with a wide range of technical and subject expertise, tackle varied client challenges from discovery to delivery. We will support your career development with ongoing training and opportunities to deepen both your technical and consulting skills. Right to Work, Location & Sponsorship We welcome applications from candidates based anywhere in the world. For roles located in the UK, Belgium, Spain, or Germany, applicants must: Hold the right to work in the country where the role is based, or be eligible for visa sponsorship, which we may consider on a case‑by‑case basis, and Be physically present in that country by their first day of employment (or able to relocate beforehand). Unless stated otherwise, relocation support is not automatically provided. Description ICF supports governments and organisations on challenges in areas such as energy, the environment, aviation, health and public policy. Our data science team works with subject matter experts to turn evidence and data into practical solutions for clients in the UK and EU. What you will do Work with data scientists and subject matter experts to understand project questions, available data and appropriate outputs. Clean, combine and explore datasets, write clear Python and document your assumptions and checks. Build and compare straightforward statistical or machine learning models, use suitable baselines and help evaluate accuracy, limitations and potential bias. Contribute to prototypes using methods suited to the problem, including language models where appropriate, and test outputs carefully before sharing them. Prepare charts, concise written findings and presentation material, explain your work to colleagues and, with support, to clients and subject matter experts. Use feedback and code review to improve your work, and contribute to reproducible project files and team knowledge sharing. What you will bring Hands‑on experience applying data science through a project, research, internship or commercial work. Experience using Python gained through employment, an internship, research or substantial personal or academic projects. A grounding in statistics and common machine learning methods, such as regression or classification, and experience of training, testing and model evaluat