MUST HAVE: Demonstrated experience building LLM-based or agentic AI systems, including tool use, retrieval, structured workflows, and rigorous evaluation. Experience with biomedical or clinical data is strongly preferred.
At M31 Biomedical AI, we are developing foundation models and intelligent systems that can reason across complex biomedical and clinical data. Building on our work in medical imaging and multimodal AI, we are expanding into agentic AI: systems that can retrieve evidence, use tools, coordinate multi-step workflows, and support scientific and clinical reasoning.
We’re seeking a full-time AI Scientist to design, build, and evaluate reliable AI agents for biomedical research and healthcare applications. You’ll work with AI researchers, clinicians, engineers, and research partners to develop systems that can operate across structured and unstructured clinical information while remaining reproducible, auditable, and scientifically grounded.
This position is ideal for someone who has moved beyond prompt engineering and has hands-on experience building agentic systems that solve real tasks.
What You'll Do Design and implement agentic AI systems for biomedical and clinical research, including multi-step reasoning, planning, tool use, retrieval, and workflow orchestration
Develop agents that interact with structured and unstructured data sources such as EHR-derived tables, laboratory data, clinical notes, radiology/pathology reports, and medical imaging metadata
Build and evaluate retrieval-augmented generation (RAG), structured-output, and tool-calling pipelines with appropriate grounding and provenance
Develop single-agent and multi-agent workflows where scientifically justified, and compare them against simpler baselines
Create rigorous evaluation frameworks for task success, factuality, calibration, robustness, reproducibility, latency, and failure modes
Implement safeguards for high-stakes biomedical settings, including constrained tool access, traceability, human review, and error analysis
Integrate LLMs and foundation models with APIs, databases, search/retrieval systems, and internal research tools




Collaborate closely with clinicians, data scientists, and engineers to translate research questions into technically sound agent workflows
Run controlled experiments and ablations to determine when agentic approaches add value over conventional ML, retrieval, or deterministic pipelines
Document and maintain reproducible workflows using Git, Python, containers, and cloud-based tools
Contribute to publications, internal reports, technical documentation, and presentations
Why Join Us Be part of a biomedical AI organization developing foundation models and advanced computational methods for healthcare and life sciences
Collaborate with leading academic, hospital, and technology research teams on high-impact biomedical research
Work on technically ambitious problems grounded in real scientific and clinical needs
Contribute to reproducible research with opportunities for publications, presentations, and recognition
Work in a mission-driven environment focused on translating computational advances into better biomedical understanding and patient impact
Required Skills & Background Master’s or PhD (or equivalent demonstrated research/engineering experience) in Computer Science, Machine Learning, Biomedical Engineering, Computational Biology, or a related field
Strong Python programming and familiarity with deep learning frameworks (e.g., PyTorch, MONAI, Transformers)
Hands-on experience building LLM or agentic systems using up-to-date model APIs and/or open-source models
Demonstrated experience with several of the following: tool/function calling; RAG and vector/semantic retrieval; agent orchestration; structured outputs; long-context systems; memory/state management; model evaluation; prompt/program optimization




Experience designing quantitative evaluations rather than relying only on qualitative demonstrations
Understanding of at least one of the following domains: Genomics or Transcriptomics
Multi-modal data integration or representation learning

Experience with data management, reproducibility, testing, and collaborative code development
Excellent problem-solving, scientific communication, and teamwork skills
Nice-To-Have Experience with biomedical, clinical, or life-science data, especially EHR or multimodal patient data
Experience with clinical NLP, medical reasoning, medical imaging, or biomedical foundation models
Experience with agent frameworks or orchestration libraries such as LangGraph, AutoGen, Semantic Kernel, or equivalent systems
Experience with model serving, inference optimization, containers, cloud/HPC environments, or distributed systems
Experience with privacy-preserving workflows, healthcare data governance, or human-in-the-loop evaluation
Strong publication record in AI/ML, NLP, agents, biomedical AI, or related areas
Experience with agentic coding tools such as Claude Code or Codex
Applications Requirements Cover letter describing a specific agentic or LLM system you built, your role in it, how you evaluated it, and why you are interested in biomedical agentic AI
GitHub portfolio, publications, preprints, or technical project links (strongly encouraged)
About M31 M31 Biomedical AI is a biomedical imaging company developing foundation models for medical image segmentation and analysis. Our technology enables universal understanding of medical images across modalities and institutions.We’re now collaborating with leading research partners to extend this vision beyond imaging – integrating multi-omics data and intelligent systems that can support biomedical discovery and patient-centered research.
Job Type: Full-time (12-month renewable contract)
Location: Hybrid remote – Toronto, ON (M5S 1A8)
Compensation: CA$65.00 – $90.00/hour, based on experience
Work-from-home option
Mentorship and publication opportunities

#J-18808-Ljbffr

📌 AI Scientist (Ontario)
🏢 M31 AI
📍 Ontario