Sr. Machine Learning Researcher

View full description

Requirements:

• Proven experience training and fine-tuning large language models, with expertise in model architecture, optimization techniques, and performance evaluation.
• Experience with modeling in the healthcare domain, clinical natural language understanding, and healthcare data and data standards (e.g., EDI, FHIR) is strongly preferred.
• Ph.D. or equivalent industry experience in fields related to machine learning, natural language processing, computer vision, computer science, electrical engineering, statistics, mathematics, optimization, or data science, plus 2-3 years of experience.
• Strong programming skills in Python, with proficiency in deep learning frameworks and tools such as PyTorch, TensorFlow, PyTorch Lightning, Hugging Face Transformers, and Kubernetes/Kubeflow, as well as experience with cloud platforms (AWS, GCP) and multi-GPU environments.
• A track record of translating research into real-world applications, with experience in rapid prototyping, experimentation, and writing production-ready code.

Responsibilities:

• Drive Applied Research: Lead the design, training, and evaluation of large language models to solve healthcare-specific challenges, advancing the state of the art in clinical Natural Language Understanding.
• Leverage Human-in-the-Loop Feedback: Work closely with cross-functional teams to integrate Human-in-the-Loop data, using it to guide model improvements and explore new methods for optimizing performance.
• Collaborate Across Teams: Partner with healthcare experts and other stakeholders to integrate qualitative insights, ensuring models align with real-world needs and deliver meaningful results.
• Stay on the Cutting Edge: Regularly evaluate advancements in ML to determine their relevance to our work, maintaining AKASA’s leading edge in responsible, high-impact healthcare AI.
• Contribute to Broader Impact: Publish and share research findings in the broader AI community, helping to advance healthcare applications of AI through peer-reviewed publications.