Senior Software Engineer (AI/ML)

Remote (UK-based) | Full-time | Flexible working | salary £82,409 + benefits including 4.5-day week and 11% employer pension

Do your best work, for the right reasons.

We're looking for a Senior Software Engineer (AI/ML) to help us build beautifully simple, high-impact digital products that support teachers and improve outcomes for pupils across the UK.

Oak is a fully remote, mission-driven organisation offering high levels of flexibility, autonomy, and purpose. We're a national not-for-profit working in partnership with teachers to create the highest-quality, sequenced curriculum and lesson resources for pupils across all subjects and age groups.

Our culture has been independently recognised through:

• Flexa verified (93% overall score, including 95% for working hours and 97% for role modelling)

• Escape the City's Top 1% Employers – based on anonymous colleague reviews of culture, development, and impact

• Investors in People Gold - through external accreditation and colleague feedback

💡 About the Role

We are looking for an experienced Senior Software Engineer (AI/ML) to join our AI Platform team.

Oak's AI Platform team builds and maintains our core AI infrastructure, including shared services, prompt libraries and reusable pipelines. This is essential in enabling our product squads to deliver safe, effective and scalable AI features that support teachers and pupils across the UK.

You'll take a leading role in shaping Oak's AI Platform, contributing towards the technical direction for reliable, safe and scalable AI capabilities. You'll work with product squads to turn ambiguous problems into robust solutions, provide technical guidance and mentorship to engineers, and help establish the patterns, tooling and standards that enable responsible AI delivery across Oak.

You'll work across the boundary between machine learning and software engineering: taking the same ownership of production reliability, testing and developer experience as other senior engineers on the team, while keeping safety and responsible AI at the centre of what you build.

In this role, you may work on:

• Enhancing Retrieval-Augmented Generation (RAG) pipelines, including context chunking, prompt construction and implementing strategies to improve relevance and accuracy.

• Optimising embedding and vector similarity systems through semantic search, clustering, filtering logic and dimensionality reduction techniques.

• Supporting model evaluation and safety tooling, such as automated test harnesses, hallucination detection, content moderation and input guardrails.

• Contributing to model fine-tuning, prompting and deployment, with a focus on balancing performance, accuracy and cost-effectiveness.

🔧 What You'll Be Doing

• Lead the continuous improvement of Oak's machine learning models and systems, from prototype through to production at scale.

• Develop and maintain highly performant ed-tech applications that are secure, scalable, highly available and accessible to the widest number of users, meeting Government Digital Service standards.

• Be an expert authority within the engineering team on the architecture, code base and AI/ML best practices, providing mentorship to less experienced colleagues.

• Design and coordinate the integration of AI/ML models with existing Oak systems, ensuring robust model serving and end-to-end testing.

• Take responsibility for the operation and reliability of production services you build — monitoring performance, resolving issues, and addressing capacity constraints before they affect users.

• Lead the implementation of new processes or tools that improve the user or developer experience.

• As a member of the Oak team, contribute to the planning and culture of the organisation.

• Work in cross-functional and product-oriented squads with colleagues from across the organisation, as required.

• Deputise for more senior colleagues and take on other general responsibilities as required.

🌱 What We're Looking For

• 5+ years' experience leading the continuous improvement of machine learning models and systems in production, alongside significant experience architecting and building high-performing web applications with modern JavaScript libraries and tools.

• Extensive knowledge of machine learning frameworks and cloud-based machine learning tools, with experience moving models from development to production.

• Understanding of machine learning ethics, fairness frameworks and responsible AI principles, including bias detection and privacy-preserving data practices.

• Experience building evaluation frameworks and automated test harnesses for AI/ML systems from scratch. Including techniques for detecting hallucinations, measuring output quality, and monitoring model performance and safety in production.

• Proven experience proposing and implementing new processes or tools that improve developer experience and lead to higher quality products.

• Experience collaborating with product teams to lead the development of high-quality, high-impact products, with great communication and collaborative skills.

💛 Our Benefits

• 25 days annual leave, plus one extra day for each year of service (up to 28)

• Additional Oak closure days over Christmas/New Year

• 11% employer pension contribution (with no minimum employee contribution)

• A 36-hour working week, with half-days on Fridays or every other Friday off

• Fully remote working — we'll support your home set-up and offer coworking options if preferred

• Twice-yearly in-person offsites to collaborate, connect, and have fun

• A culture that genuinely supports flexibility, autonomy, and trust

🌈 Inclusion and Belonging

We believe diverse teams build better products. We warmly welcome applicants from all backgrounds, particularly those who are underrepresented in the tech and education sectors.

If you're from an underrepresented group, there's a good chance you're discounting yourself before you've even started. That's more common than you'd think, and it means we may miss out on brilliant people. If you're excited by this role but don't meet every requirement, please apply anyway.

We use the Applied platform to help reduce bias in our hiring process. Answers are anonymised and reviewed by a panel of humans.

📌 Key Info

• Location: Remote, but you must be based in the UK with the legal right to work here

• Sponsorship: Unfortunately, we're unable to offer visa sponsorship at this time

• Closing date: We'll be reviewing applications as they come in and may close the role early

If this sounds like the kind of role and team where you could do your life's best work, we'd love to hear from you.

Next steps

You'll answer a few admin questions and some questions related to your day-to-day job. Your answers will go through our sift process: all answers will be anonymised, randomised, and then reviewed by a panel of reviewers (real humans), so it's your thinking that gets assessed, not your CV.

If you're shortlisted, we'll invite you to interview across three stages. We're aiming to start interviews the week beginning 17th August 2026.

Stage 1: An introductory call to learn more about your skills and experience, and a chance for you to ask us about the role. This takes about 30 minutes.

Stage 2: A technical interview over Zoom, combining a hands-on pairing session or technical walk-through, followed by some technical interview questions. This takes around two hours.

Stage 3: A collaboration interview, focused on how you work with both engineering and non-engineering colleagues. This takes about 1 hour.

We love giving feedback, so at the end of the application process, we'll share our assessment of your performance.

We are receiving excellent responses to our job advertisements. This may lead us to close the role early, so if you are considering applying, don't leave it too long to avoid missing out.

We are an equal opportunity employer.

We are committed to a policy of Equal Employment Opportunity and are determined to ensure that no applicant or employee receives less favourable treatment on the grounds of gender, age, disability, religion, belief, sexual orientation, marital status, or race, or is disadvantaged by conditions or requirements which cannot be shown to be justifiable.