Senior AI Scientist in London

Reporting To: Director of Data Science, AI & Research

Location: Hybrid, working from home and our office based in West London

Salary: £55,000 - £75,000 per annum, dependant on experience. We aim to be transparent about remuneration at MHI. As a charitable organisation, salaries for this role are predetermined and not negotiable. Please consider the advertised salary before applying.

Hours of Work: 40 hours per week, inclusive of a daily 1 hour paid lunch break.

Contract: Permanent

Benefits:

• 25 days annual leave per year, plus public holidays;
• Company electronic devices;
• Enhanced salary sacrifice pension scheme;
• Private health insurance after completion of probation;
• Eligibility for a Blue Light discount card.

Closing Date: 20th November 2026. We will be interviewing on a rolling basis and reserve the right to close the job advert early if we receive a high number of applicants, so we encourage applying early.

Job Purpose: To use AI and data science to understand and improve MHI’s services, and to evaluate AI tools so they are used safely, responsibly and only where they add value.

Impact: Our Data Insights work directly shapes our services and drives our innovation pipeline designed to develop new products and solutions. For example, our recent work, in collaboration with Imperial Global Health Institute, showed an increase in 13s and under using the Shout service, which lead to the development of a proof-of-concept early intervention specifically for 10-13 year olds.

Team: You’ll join the Data Insights team, working within a multidisciplinary team (~6-10 people) including data engineers, data scientists, data analysts, research psychologists and a policy lead, with close collaboration with our technical and security teams and links to our clinical and service delivery colleagues.

Data and tools: You’ll work with a unique dataset: anonymised text from more than 4 million Shout conversations, plus a growing dataset from The Mix’s services, including peer support and counselling. All data is held in a secure AWS environment, with Amazon SageMaker for model development and Amazon Bedrock for access to frontier AI models.

Responsible data use: Data are securely governed with access on a strict needs basis and research ethics sought where projects extend beyond research purposes for public interest as outlined in the data privacy policy.

Wellbeing: As a mental health service that supports those in crisis, we provide insights to colleagues working with sensitive and potentially distressing content. Wellbeing is important to us, and we provide private healthcare and an EAP service that offers mental health support. We also have a number of mental health first aiders that can provide confidential support.

Key Responsibilities:

• Conduct research: Investigate open questions about our services and the people who use them, designing studies that produce robust evidence with actionable reports.
• Analyse large-scale text: Apply ML and NLP methods to large volumes of free-text data, such as conversations from our crisis, peer support and counselling services, with careful attention to validity, reliability and bias.
• Prototype and test: Build technical prototypes to test whether AI approaches are viable, and work with colleagues to take promising approaches forward.
• Evaluate AI rigorously: Design and run evaluations of AI models and tools, including prototypes and pilots developed across MHI, validating automated methods against clinical expert judgement and safety testing for high-risk conversations.
• Promote responsible AI: Assess risks, including bias, privacy, safety and potential harm, and help develop appropriate safeguards.
• Shape strategy: Work with the Director of Data Science, AI & Research to develop MHI's AI strategy and research priorities within the Data Insights team.
• Contribute to technical leadership: Help shape technical and methodological standards, support and mentor colleagues, and advise on AI capabilities, limitations, opportunities and risks.
• Collaborate and share knowledge: Work with clinical, service delivery and external partners, communicate complex ideas clearly to non-technical audiences, and contribute to reports, publications and presentations.

Person Specification:

Essential:

• A PhD or equivalent experience in a quantitative field, such as computer science, machine learning, statistics, physics, computational neuroscience or quantitative psychology, and a track record of high-quality, well-tested and reproducible technical work you have delivered yourself.
• Strong Python skills, including proficiency with modern deep learning frameworks (e.g., PyTorch, TensorFlow, JAX, Hugging Face), and working knowledge of SQL for querying data.
• Hands-on experience applying machine learning and NLP, including large language models, to real-world free text, and turning the results into clear, practical findings.
• A strong grounding in statistics and research design, and experience designing and running evaluations of AI models or systems, including validation against human judgement.
• Sound judgement about when AI is and is not the right approach, a preference for the simplest method that solves the problem, and an understanding of responsible AI, including fairness, transparency, privacy and safety.
• Proactive and self-motivated, with the ability to take initiative, work independently and manage your own priorities, including spotting problems worth solving.
• Excellent communication and collaboration skills, including explaining technical work to non-specialists and supporting or mentoring colleagues.
• Ability to handle sensitive mental health data with care, and a commitment to using technology ethically to create positive social impact.

Desirable:

• Experience applying AI in mental health, healthcare, or other social and public-interest settings.
• Knowledge of AI governance, assurance or risk management, including regulation relevant to AI in health.
• A track record of publications or research reports in machine learning, NLP, AI evaluation or digital mental health.
• Experience working in AWS or another cloud platform, including running machine learning workloads and using managed AI services.
• Experience working with external research partners or managing collaborative projects.
• Experience translating research into practical tools, recommendations or policy.

This role is subject to eligibility to work in the UK, plus satisfactory background and reference checks. Unfortunately, we are not able to offer visa sponsorship.

We understand that AI is increasingly part of everyday life, and you might choose to use it when putting together your application. While AI can be a helpful tool, we ask that anything you submit reflects your own skills, experiences, and perspective. We value authenticity and integrity, and we want to see what you uniquely bring to our team and why our mission resonates with you personally.