Company Description

Karyon Bio is a precision-liver-health company uniting scientists, clinicians, and AI engineers to transform the early detection and management of fatty liver disease (MASLD/MASH). We work at the intersection of life sciences, multimodal AI, and healthcare systems integration.

Our vision is to build fusion models that combine foundation models across multiple modalities—imaging, omics, clinical/EHR, and real-world data—and orchestrate them using Agentic AI to deliver clinically actionable insights. By connecting hospital systems, payers, and advanced AI, we aim to bring earlier diagnosis, better risk stratification, and personalized care to millions of patients worldwide.

Join a mission-driven, forward-looking team where you will work on real-world problems, not toy datasets—and see your work move toward impact in hospitals and health systems.

Role Description

This is a remote internship role for an Artificial Intelligence Engineer at Karyon Bio. You will help design and implement fusion AI architectures that can:

• Combine outputs from multiple foundation models (vision, language, tabular, and omics) into unified decision pipelines.
• Leverage Agentic AI (tool-using, multi-step reasoning agents) to orchestrate data ingestion, preprocessing, inference, and reporting across different healthcare data sources.

You will:

• Contribute to the design and prototyping of multimodal pipelines integrating:
• Medical imaging models (e.g., ultrasound, CT, MRI)
• Text / NLP models for clinical notes, guidelines, and literature
• Tabular models for lab values, demographics, and claims/EHR data
• Implement and test agent workflows that can call different models/tools, reason over their outputs, and generate clinician- or patient-facing summaries.
• Develop reusable software components, APIs, and utilities that make it easier to plug new foundation models into our fusion framework.
• Work closely with data scientists, clinicians, and product teams to translate clinical questions into AI workflows and evaluation metrics.

This role is ideal for someone who wants hands-on experience at the frontier of multimodal foundation models + Agentic AI in healthcare and is excited to learn fast in a high-impact startup environment.

Qualifications

Core Technical Skills

• Solid understanding of machine learning and deep learning, especially:
• Pattern recognition and neural networks
• Experience with at least one of: vision models (CNNs/Vision Transformers), NLP models (Transformers/LLMs), or tabular models.
• Strong programming skills in Python and experience with ML frameworks such as PyTorch or TensorFlow.
• Familiarity with:
• Working with foundation models (e.g., open-source LLMs, vision models)
• Building data pipelines and APIs
• Basic MLOps concepts (versioning, reproducibility, experiments tracking) is a plus.

Agentic & Multimodal Mindset

• Interest or experience in Agentic AI (tool-using agents, function calling, multi-step reasoning workflows).
• Curiosity about multimodal learning —combining text, images, and structured data into a single decision system.
• Ability to think in terms of systems rather than single models (how data flows, how models interact, how results are consumed).

Domain & Soft Skills

• Background in Computer Science, AI/ML, Data Science, or related field.
• Ability to work collaboratively in a distributed, multidisciplinary team (engineering, clinical, business).
• Strong written and verbal communication skills to explain complex technical ideas in a clear way.
• Experience or interest in healthcare, life sciences, or digital health is a plus—but a strong willingness to learn is even more important.
• If you are excited about building AI systems that go beyond single models—systems where agents coordinate multiple foundation models across modalities to solve real clinical problems —we’d love to hear from you.