Principal AI Engineer

As the Principal AI Engineer, you will lead the technical vision and architecture for the agentic AI system reimagining CDI. You will oversee the design of autonomous AI agents that can independently analyze medical records, detect inconsistencies, and simulate queries or updates in EHRs. This role requires expertise in AI orchestration, multi-agent systems, and healthcare domain knowledge to ensure the system operates reliably and ethically.

Key Responsibilities:

• Architect and implement agentic AI frameworks using technologies like LangChain, CrewAI, or custom multi-agent systems to enable autonomous decision-making in CDI processes, incorporating MCP for model context management and A2A protocols for inter-agent communication.
• Collaborate with data scientists and engineers to integrate machine learning models for natural language processing (NLP) of clinical notes, ICD-10 code prediction, and severity assessment, including finetuning transformers-based models and SLMs for optimized performance.
• Ensure AI agents comply with healthcare regulations (e.g., HIPAA, CMS guidelines) and incorporate ethical AI principles, such as bias mitigation and explainability, while leveraging knowledge graphs for semantic understanding of clinical data.
• Lead pod-level sprints, mentor junior team members, and coordinate with the second pod for system scalability and interoperability, focusing on custom agentic architectures and optimizations for model costs and latency.
• Prototype and deploy AI agents in simulated and real-world EHR environments, monitoring performance metrics like accuracy in documentation improvement and reduction in manual interventions.

Qualifications:

• Master's or PhD in Computer Science, AI, or a related field.
• 8+ years of experience in AI engineering, with at least 3 years in agentic or autonomous AI systems, including hands-on work with finetuning models, SLMs, knowledge graphs, transformers architecture, closed-source (e.g., GPT series) and open-source models (e.g., Llama, BERT), custom agentic architectures, optimizing model costs and latency, and implementing MCP and A2A protocols.
• Proficiency in Python, TensorFlow/PyTorch, NLP libraries (e.g., spaCy, Hugging Face Transformers), and cloud platforms (e.g., AWS, Azure).
• Experience in healthcare AI, preferably with EHR integrations (e.g., Epic, Cerner) and knowledge of medical ontologies (e.g., SNOMED CT).
• Strong leadership skills, with a track record of delivering complex AI projects on time