Machine Learning Engineer

Limbic's vision is to ensure the highest quality therapy is accessible to everyone, everywhere. The way we do that is by deploying AI responsibly, using it to augment clinical care and reduce barriers to accessing therapy at scale. Limbic is already used in over 40% of Talking Therapies in the NHS, which makes us the largest-ever deployment of generative AI for direct patient care. Limbic is deemed safe and effective after successfully gaining UKCA Class IIa certification, the only AI mental health chatbot to do so. Every day, 1K+ patients interact with Limbic, making therapy more accessible. Limbic goes beyond replacing existing processes with AI and helps 189 people recover from their mental health challenges who otherwise would not have.

We're seeking a Machine Learning Engineer to help deliver best-in-class AI for mental health care. You will develop, refine, and maintain the functionalities that allow us to serve our ML-capabilities in the product. You will play a key role in translating our AI research into products that help our users in the real-world.

What you’ll do:

• Design, build and deploy ML powered features end to end, including contributing to the architecture of systems combining LLM features with our in-house safety guardrails and clinical ML-models.
• This includes working with leading commercial AI models from OpenAI, Anthropic or Google, and also working with in-house self-hosted instances of models like Llama.
• Implementing and refining evaluation methodologies combining user testing, LLM evals, and clinical trials.
• Implementing continuous monitoring strategies for production deployments.
• Collaborate with team members through code review and contribute to engineering best practices.
• Collaborate with application engineers at the interface of application and ML engineering, integrating with the rest of the product teams work (e.g. security, orchestration, data).
• Help build and tune custom models.

Sample recent projects:

• Architect and deploy a voice agent for mental health service phone calls, integrating safety guardrails and implementing continuous monitoring and improvement systems.
• Collaborate with clinicians to develop an AI triage system that conducts text-based patient intake conversations, gathering critical information to provide clinicians with actionable insights before appointments.
• Train, evaluate and deploy models to detect problematic user interactions.
• Build and run experiments comparing the therapeutic sessions powered by voice vs text based AI agents.

We encourage women and individuals from diverse backgrounds to apply and join our team. We believe in creating an inclusive and supportive work environment where everyone can contribute their best.

Requirements:

• 2+ years of industry Machine Learning experience.
• This position requires availability during standard business hours for our teams across Europe (UTC to UTC+3).
• Ability to work from London (Spitafields) office 1-day per week is preferred.
• Limbic cannot provide visa sponsorship.
• Experience working with LLM features.
• Strong Python ability.
• Experience with at least 1 backend python framework (e.g. we use FastAPI).
• Solid understanding of statistics in the context of ML model evaluation and user activity analytics.
• Experience with databases (SQL or NoSQL).
• Familiarity with cloud platforms (AWS, GCP, Azure) and containerization technologies (Docker).
• Proficiency with version control systems, preferably Git.
• Interest in the mental healthcare space.
• Experience with Typescript / Javascript is a bonus.

Benefits:

• An amazing office in central London (flexibility regarding working from the office and wfh).
• 25 days PTO.
• Pension scheme.
• Enhanced parental leave packages (UK).
• Equity share options.
• Support with purchasing work-related books and materials.

We take employee wellbeing seriously at Limbic and in addition to the above we offer:

• Quarterly life days (Enjoy 4 paid days off per year to use whenever you choose - one each quarter - to relax, recharge, or take care of personal matters).
• Access to mental health support.