About Us

At Prenuvo, we are on a mission to flip the paradigm from reactive “sick-care” to proactive health care. Our award-winning whole body scan is fast (under 1 hour), safe (MRI has no ionizing radiation), and non-invasive (no contrast). Our unique integrated stack of optimized hardware, software, and increasingly AI, coupled with the patient-centric experience across our domestic and global clinics, have allowed us to lead the change against “we caught it too late again".

We are looking for a Senior AI Applied Scientist II (Agentic) to independently own a model domain end-to-end within our AI Research org, from problem framing through validated, production-deployed models.

As Prenuvo’s AI org moves from task-specific models toward foundational work (meaning pre-trained backbones, multi-task heads, and the addition of clinical text and structured data), we need scientists who design experiments rather than only run them, diagnose failure modes on their own, and shape how the team approaches a whole class of problems. You will work at the intersection of deep learning, medical imaging, and clinical translation, accountable for both the rigor of the science and the delivery of the work.

This hybrid role is based in Vancouver. We expect you to be technically credible enough to review model architectures, challenge assumptions, and roll up your sleeves when it matters. We also expect you to build a team environment where people do their best work.

This role does not provide visa sponsorship. Applicants must be legally authorized to work in Canada at the time of hire and must not require employer sponsorship for a work visa (current or future).

Help reshape the world through proactive healthcare while working with cutting-edge technology and high performing teams with deep expertise - join us to make a difference in people’s lives!

What You’ll Do

LLM Fine-Tuning & Domain Adaptation: Lead the end-to-end fine-tuning of large language models (SFT, PEFT/LoRA, DPO/RLHF) on clinical text, radiology reports, and longitudinal patient histories to build specialized, domain-adapted clinical models.
Agentic AI System Architecture: Design, develop, and operationalize autonomous and semi-autonomous AI agent systems utilizing multi-step reasoning, tool use, workflow orchestration, planning, and memory architectures to automate and accelerate complex research and clinical workflows.
Comprehensive NLP & LLM Evaluation: Establish robust, automated, and human-in-the-loop NLP evaluation frameworks to benchmark clinical accuracy, detect hallucinations, assess reasoning capabilities, and ensure safety, reliability, and regulatory compliance of deployed language models.
Multimodal Integration: Collaborate on multimodal systems connecting imaging foundation models with LLMs, structured lab data, and clinical text to create holistic patient insight platforms.
Experiment Design & Validation: Own experiment design and validation study scoping for your domain, diagnosing failure modes independently and establishing rigorous scientific benchmarks.
Data Quality & Clinical Framing: Drive data quality standards, clinical annotation criteria, and knowledge grounding (e.g., RAG, ontology alignment) for language and agentic pipelines in close collaboration with medical experts.
Technical Leadership & Collaboration: Partner across clinical, annotation, and ML Engineering teams, raising the code-quality bar and contributing reusable architectures and frameworks adopted org-wide.
Production Delivery: Deliver production-ready LLM and agentic models with documented validation, clinical sign-off, and continuous monitoring benchmarks, owning the technical narrative for your domain.

What You’ll Bring

Technical depth

Strong MSc or PhD from a top-tier institution in CS, computational linguistics, AI/ML, statistics, biomedical engineering, or a related field.
A minimum of 4 academic or industry years of ML/NLP experience, with a strong publication or applied-research record including first-author works in top ML/NLP venues (e.g., NeurIPS, ACL, EMNLP, NAACL, ICML, ICLR).