Acentra Health is looking to innovate health solutions that deliver maximum value and impact by building, optimizing, and integrating machine learning (ML) and large language model (LLM) solutions to drive innovation, improve healthcare outcomes, and create operational efficiencies.
Requirements
- Proficiency in Python and ML/AI frameworks (TensorFlow, PyTorch, Hugging Face, scikit-learn)
- Hands-on experience with LLMs, embeddings, vector DBs, and RAG pipelines
- Cloud experience with AWS and Azure for deploying and managing AI systems
- Familiarity with MLOps/LLMOps and CI/CD practices
- Experience deploying ML/LLM systems in healthcare or other regulated environments
- Experience with LangChain, LlamaIndex, or agentic AI frameworks (AutoGen, CrewAI, Semantic Kernel)
- Knowledge of healthcare data standards (HL7, FHIR) and compliance (HIPAA)
Responsibilities
- Design, develop, and deploy LLM-powered applications (e.g. summarization, intelligent assistants, document processing, automation)
- Strong skills in API development and system integration, with the ability to contribute across the full stack
- Build and optimize RAG pipelines using embeddings and vector databases (e.g., Pinecone, Weaviate, FAISS)
- Support ML pipelines including data preparation, training workflows, and deployment into applications
- Support and implement MLOps/LLMOps practices such as CI/CD pipelines, versioning, monitoring, retraining, and governance
- Mentor and coach engineers on best practices, solution architecture, and applied AI strategies
- Partner with business and product teams to translate requirements into effective AI solutions
Other
- Position is remote, but candidates based in Raleigh/Durham/Cary, NC or DC area preferred.
- Stay current on advances in LLMs, GenAI, and compliance requirements (HIPAA, GDPR, FDA)
- Read, understand, and adhere to all corporate policies including policies related to HIPAA and its Privacy and Security Rules
- Excellent communication skills and experience mentoring and collaborating with engineers
- Exposure to multi-agent system design and implementation
