Masters Degree in a relevant field., 3+ years of experience in machine learning or NLP roles, focusing on LLMs and GenAI., Strong proficiency in Python and deep learning frameworks like PyTorch or TensorFlow., Experience with healthcare data and compliance (HIPAA/GDPR)..
Key responsabilities:
Design and implement RAG architectures using various LLMs.
Build and maintain retrieval pipelines for health data.
Integrate RAG outputs into user-facing applications for reliable responses.
Collaborate with product and data science teams to enhance model performance.
Report This Job
Help us maintain the quality of our job listings. If you find any issues with this job post, please let us know.
Select the reason you're reporting this job:
all.health is at the forefront of revolutionizing healthcare for millions of patients worldwide. Combining more than 20 years of proprietary wearable technology with clinically relevant signals, all.health connects patients and physicians like never before with continuous, data-driven dialogue. This unique position of daily directed guidance stands to redefine primary care while helping people live happier, healthier, and longer.
Education
Masters Degree
About the Role
You will design, build, and optimize RAG pipelines that combine large language models (LLMs) with domain-specific retrieval systems, enabling natural language understanding and reasoning over patient data, clinical guidelines, and health records. Your work will directly impact how patients and clinicians interact with our platform, enabling safe, accurate, and context-aware content surfacing.
Responsibilities
Design and implement RAG architectures using open-source and potentially proprietary LLMs (e.g., LLaMA, Mistral, OpenAI, Anthropic).
Build and maintain retrieval pipelines over structured and unstructured health data (EHRs, patient notes, device logs, clinical documentation).
Develop indexing strategies using vector databases (e.g., FAISS, Weaviate, Pinecone) and embedding models (e.g., BioBERT, ClinicalBERT).
Integrate RAG outputs into user-facing applications, ensuring responses are grounded, reliable, and privacy-compliant.
Work closely with product, clinical, and data science teams to fine-tune prompts, evaluate responses, and iterate on model performance.
Build evaluation pipelines for factuality, relevance, and safety using synthetic and real-world datasets.