Senior Software Engineer - AI Systems

## About the role ### Why We Exist Nurse triage is the first touchpoint for most pediatric care—and it's overwhelmed. High call volume, manual workflows, incomplete documentation, and misaligned reimbursement leave clinicians exhausted and patients waiting. Lucea Health is building an AI-native triage platform that: - Handles patient intake autonomously via AI voice + text - Assists nurses using evidence-backed protocols - Extracts structured clinical data from unstructured conversations - Generates EHR-ready documentation automatically - Hands off to nurses with full context when human judgment is needed We're not replacing nurses. We're removing the administrative burden that pulls them away from care. We're live with our first design partner, have validated that automation (not just augmentation) is where the value lives, and are now focused on making our AI workflows production-ready for active use by clinicians. ### Why This Role Matters Our core product is AI. The quality of our LLM outputs—transcripts, data extraction, clinical summaries, autonomous agent conversations—is the difference between a product nurses trust and one they don't. We're looking for a senior engineer to own AI quality and reliability across our platform. You'll ship production code, tune prompts, build eval frameworks, and fix the edge cases that matter in healthcare. This is a compliance-critical environment; HIPAA, encryption, and auditability aren't optional. As a small team with a technical, product-focused CEO, you'll have direct line of sight to product strategy and direction. This is a hands-on building role, not a research role. ## What You'll Own ### AI Quality and Reliability - Own prompt engineering across our LLM workflows (transcription, data extraction, note generation, autonomous agent) - Build evaluation and testing frameworks for non-deterministic outputs - Identify and fix hallucination risks, extraction failures, and edge cases - Design guardrails, fallbacks, and confidence thresholds for clinical safety - Instrument and monitor AI system performance in production ### Production LLM Systems - Design and improve our LLM orchestration (multi-step workflows, context management, cost/latency tradeoffs) - Work with real-time voice and streaming data (Twilio, Anthropic, OpenAI, Livekit, transcription pipelines) - Build infrastructure for prompt versioning, A/B testing, and iteration - Optimize for reliability, latency, and cost at scale ### Backend & Platform - Build and maintain core backend systems (Python, Node.js, PostgreSQL, Redis) - Evolve our data models for clinical information extraction - Shape API design and system architecture decisions as we scale ## What We're Looking For ### Must Have - 5+ years building production software, including backend systems - AI-native - you use AI coding tools daily and ship faster because of it - Hands-on experience building and shipping LLM-powered features (not just prototypes) - Strong intuition for prompt engineering and working with foundation models - Experience