Apply on Kit Job: kitjob.ca/job/2g9aj1

Role Overview

You'll own the intelligence layer that makes this system clinically safe: the LLM orchestration pipeline, safety systems, and memory and retrieval architecture. You work directly with our CTO and CEO/clinical founder. This is a full-time, senior IC role with real technical authority from day one. What You'll Own

LLM orchestration: prompt assembly, context management, and multi-model routing Safety and compliance layer for clinical AI interactions Memory and retrieval architecture (vector search, RAG pipelines) Evaluation frameworks and observability infrastructure to ensure reliable production performance Must-Have Experience

Built and shipped LLM/NLP systems on clinical or health data in production — not just prototypes or academic environments Designed and owned a RAG pipeline end-to-end in production — not just integrated a library Built content safety systems for LLM outputs beyond keyword filtering; understands failure modes in high-stakes or sensitive contexts Instrumented a production LLM system with evaluation and observability infrastructure — can reason about model performance and degradation Hands-on production experience with at least one LLM orchestration framework (e.g. LangChain, LangGraph), a vector database, and cloud LLM infrastructure Robust Plus

Experience building AI systems handling PHI under HIPAA, PIPEDA, or equivalent healthcare privacy frameworks Experience designing human-in-the-loop systems for clinical workflows Published research or applied work in healthcare AI Why Join

Build the clinical AI system, not a wrapper — this is the product, not a feature Work directly with a psychiatrist who understands both the clinical nuance and the engineering constraints Competitive salary and meaningful equity, commensurate with experience A chance to define how safe, human-aligned AI shows up in mental health

#J-18808-Ljbffr

Apply on Kit Job: kitjob.ca/job/2g9aj1

📌 Principal AI Engineer (Toronto)
🏢 GrowthSpan
📍 Toronto