Get AI-powered advice on this job and more exclusive features.
Direct message the job poster from Cerebria.ai
Cerebria is at the forefront of AI-driven healthcare innovation, developing advanced machine learning solutions that enhance patient safety and improve cardiovascular diagnostics. Our core technology focuses on real-time coronary physiology, supporting interventional cardiologists with intelligent insights during procedures. In parallel, we are building cutting-edge education tools that leverage generative AI to deliver personalized, scalable learning experiences for healthcare professionals and patients.
Founded by practicing cardiologists, we're a small but growing startup backed by deep expertise in medical devices, regulatory compliance, and systems integration. From AI model development to seamless deployment in clinical and educational environments, we deliver end-to-end innovation. With headquarters in London, engineering operations in California, and a new engineering team being established in South Africa, our global team is united by a mission to make cardiovascular care and education smarter, safer, and more effective through intelligent technologies.
The role
We’re looking for an AI Engineer to build and integrate production-ready applications powered by Large Language Models (LLMs). You’ll work on both our regulated medical device products and AI-powered education tools, helping to build systems that deliver personalized, traceable, and clinically aligned experiences across both domains.
The ideal candidate combines technical expertise with strong product vision and a passion for collaborative problem-solving. You\'ll work closely with our team, clinicians, and domain experts to rapidly prototype AI-driven concepts, validate them in realistic settings, and scale them into production-ready applications that meet the real-world demands of clinical safety, performance, and usability.
This role is not a fit for junior candidates, or anyone without prior production experience shipping LLM systems.
This position is based in South Africa; some international travel may be required for team meetings or project collaboration. Please do not apply if you are not based in South Africa .
What you’ll do
- Develop and integrate LLM-based agentic systems for medical decision support and personalized education
- Build multi-step reasoning workflows and agents that adapt dynamically to clinical context, learning goals, and user interaction
- Collaborate with clinicians and domain experts to co-design, test, and iterate on new AI concepts
- Translate early-stage prototypes into robust, maintainable, and compliant production code
- Design RAG pipelines and knowledge orchestration systems for structured information access and traceable output generation
- Deploy and maintain LLM applications across both regulated software systems (medical devices) and non-regulated platforms (education tools) to optimize agentic workflows for inference performance, latency, and cost-efficiency
- Support the integration of AI models into regulated medical device software, with attention to performance optimization, safety, and runtime constraints
- Maintain thorough technical documentation, including model architectures, design specifications, training procedures, testing protocols, and performance benchmarks
- Adhere to and support implementation of applicable quality and regulatory standards, including IEC 62304, ISO 13485, and FDA / CE requirements
- Participate in design reviews, sprint planning, and collaborative meetings across teams
What we’re looking for
Must have experience
Bachelor’s or Master’s degree in Computer Science, Artificial Intelligence, Machine Learning or a related fieldMinimum 2–5 years of experience working in Artificial Intelligence, Machine Learning, or Software EngineeringProven hands-on experience building production LLM applications or agentic systemsDemonstrable experience with multi-agent frameworks, LLM orchestration platforms, or function-calling implementationsDemonstrable experience in prompt engineering, model evaluation, and implementing guardrails against hallucinationsVery strong Python proficiency including modern LLM frameworks (LangChain / LangGraph, LlamaIndex) and LLM provider APIsExperience with web frameworks for API development and deployment (FastAPI, Flask, Django) and RESTful service architectureProven ability to collaborate with non-technical stakeholders to turn ideas into working softwareExperience with prototyping and iterative design workflows, particularly in startup or research settingsProficient and practiced in production-grade engineering practices including testing, version control, and modular designExperience working in or adjacent to regulated environments (medical devices, diagnostics, healthcare IT)Background in educational technology or intelligent tutoring systemsExperience implementing evaluation frameworks for LLM reliability, factual accuracy, and output traceabilityKnowledge of cloud-based deployment and MLOps tools for model serving and monitoringKnowledge of regulatory frameworks for AI / ML in medical devices (FDA, CE marking processes)Understanding of healthcare data standards (HIPAA, HL7 / FHIR) and clinical workflowsKnowledge of embedded AI systems or real-time model inference in hardware-constrained environmentsBackground in safety-critical applications (medical, automotive, aerospace, manufacturing)Why join Cerebria?
Remote-first culture with global impactHelp shape both the future of interventional cardiology and clinical educationWork closely with clinicians to build real-world tools with measurable outcomesBridge the gap between early AI innovation and regulated healthcare deploymentEnd-to-end ownership from research to productionSalary R700 000 to R1 200 000 per annumReady to make AI that matters? Let\'s transform healthcare together.
Seniority level
Mid-Senior levelEmployment type
Full-timeJob function
Engineering and Information TechnologyIndustriesSoftware DevelopmentWe’re unlocking community knowledge in a new way. Experts add insights directly into each article, started with the help of AI.
#J-18808-Ljbffr