Define and build AI systems that turn glucose data and model outputs into personalized insights. Lead architecture for LLM-based agents, establish safety and evaluation standards, and productionize scalable inference services while collaborating with data science, clinical, regulatory, product, and mobile engineering teams.
Responsibilities
- Define technical vision and architecture for agentic and LLM-based features across cloud and device.
- Design agents that combine glucose history, model predictions, meals, activity, and user context to generate insights.
- Own agent orchestration, tool use, memory, and retrieval patterns, selecting models based on safety, cost, and latency.
- Establish guardrails and clinically grounded evaluation frameworks for accuracy, safety, and user impact.
- Ensure AI outputs are traceable and auditable, and partner with regulatory and quality teams on risk analysis and documentation.
- Productionize time-series and personalization models and build monitored batch and real-time inference services.
Required Skills
- Large language models
- Agentic AI systems
- AI system architecture
- Agent orchestration
- Tool use and retrieval patterns
- Model evaluation frameworks
- Time-series modeling
- Personalization modeling
- Batch and real-time inference
- Cloud-native production services
