2 months ago
London, United KingdomSenior
Responsibilities
- Integrate an on-device LLM with live neuromonitoring signals and build the clinician-facing query and response layer.
- Develop model evaluation, prompting, and guardrail approaches to ensure safe and consistent clinical outputs.
- Optimize inference latency, memory usage, and throughput on constrained on-device hardware.
- Design and develop medical-device software across real-time device behavior and clinician-facing interfaces in accordance with IEC 62304 and ISO 14971.
- Ensure AI deployments meet cybersecurity requirements, regulatory standards, and FDA guidance for AI/ML-enabled devices.
- Maintain fleet reliability and collaborate with hardware, product, and clinical teams to diagnose and resolve issues.
Requirements
- 7+ years of hands-on software engineering experience.
- Professional experience with C++ and Python.
- Practical experience with LLM prompting, guardrails, retrieval-augmented generation, evaluation, and model behavior analysis.
- Experience building or extending a model and integrating it meaningfully into a user-facing production product.
- Experience integrating third-party libraries or complex components into larger, performance-sensitive, or resource-constrained systems.
- Ability to reason about latency, memory, and throughput on constrained hardware.
- Strong experience with Git and CI/CD.
- Preferred: edge or on-device AI experience with platforms such as Jetson, Android, iPad, Raspberry Pi, or microcontrollers.
- Preferred: hands-on experience with llama.cpp or similar tools and evaluation of small on-device models such as Qwen, Gemma, MedGemma, or custom models.
- Preferred: hardware exposure, AWS cloud experience for LLMs, or experience in a regulated or safety-critical industry.
Benefits
- Company equity plan.
- Salary-sacrifice pension scheme.
- Private medical, dental, and vision insurance with medical history disregarded.
- Group life assurance at 4x annual income.
- Comprehensive mental health support with unlimited one-to-one sessions.
- Unlimited holiday allowance plus bank holidays and one week of remote working per quarter.
- At least four days per week in the Kings Cross office, with one flexible work-from-home day.
- Daily £10 JustEat lunch voucher and free dinner when working late.
- Fresh fruit deliveries twice weekly plus snacks and drinks.
- YuLife subscription with discounts linked to daily steps and meditation.
Categories
About CoMind
CoMind builds non-invasive brain-monitoring hardware and AI software for hospital clinicians, using light-based neuroimaging to capture multidimensional brain activity. The company designs and manufactures its device and the CoVision analytics platform, supporting care in intensive care, surgery, and emergency settings. Privately held and founded in 2018, CoMind is headquartered in London and develops regulated medical devices for FDA clearance and CE marking.