About the role

InTouchNow AI is looking for an AI Research Engineer to join InTouch Labs, our applied research group, based at our Borehamwood office. This is a hands-on role focused on turning the latest research in large language models, speech and voice AI into production capability inside our conversational AI reception platform for NHS general practice.

You will work at the boundary between research and engineering: reading and reproducing recent papers, running structured experiments against real GP practice call data, and shipping the improvements that survive contact with production. This is a rare opportunity to influence how AI is used at scale across NHS primary care, on a product handling tens of thousands of patient calls each week.

What you will do

• Design and run experiments to improve our voice AI’s understanding, response quality, transfer decisions and safety behaviours.
• Evaluate and benchmark new LLM, ASR and TTS models against our production traffic and internal quality metrics.
• Build reproducible evaluation harnesses so improvements can be measured objectively across releases.
• Prototype new capabilities (agentic workflows, retrieval-augmented generation, fine-tuning, guardrails) and validate them on real call data.
• Work with the engineering team to move validated prototypes into production and measure their real-world impact.
• Read, summarise and reproduce relevant research in LLMs, speech, dialogue systems and voice AI, and translate findings into concrete product ideas.
• Contribute to our patent portfolio by identifying novel approaches worth protecting.
• Collaborate with clinical safety, product and customer teams to ensure new AI behaviours are safe, appropriate and useful for GP practices.
• Contribute to InTouch Labs publications, blog posts and conference talks that share our applied research with the wider healthcare AI community.

What we are looking for

• At least 5 years’ hands-on experience applying machine learning or AI research in production settings.
• Deep understanding of large language models, prompt design, evaluation and their limitations.
• Practical experience with modern LLM tooling and orchestration frameworks.
• Strong Python skills, with the ability to move quickly from prototype to production-quality code.
• Experience designing and running experiments end-to-end: hypothesis, dataset, evaluation, analysis.
• Comfortable working with speech models (ASR/TTS) or the willingness to learn them quickly.
• Excellent written communication, particularly the ability to summarise research clearly for both technical and non-technical audiences.
• Willingness and ability to commute to our Borehamwood office regularly as part of a hybrid working pattern.

Nice to have

• Published research or open-source contributions in NLP, speech, dialogue systems or voice AI.
• Experience fine-tuning or evaluating large language models on domain-specific tasks.
• Experience with real-time voice AI systems, latency-sensitive inference or streaming pipelines.
• Experience working on safety-critical or regulated AI products.
• Familiarity with UK healthcare, NHS primary care or clinical safety standards.
• Experience contributing to patents, papers or technical white papers.

What you will get

• Competitive salary of £85,000 to £110,000 per year.
• Private health insurance.
• Free on-site office parking.
• 20 days’ annual leave plus bank holidays.
• Workplace pension scheme.
• Discretionary bonus.
• Opportunity to join the company share scheme.
• Hybrid working arrangement, with regular time at our Borehamwood office.
• Opportunities for professional development, conference attendance and publication.