Clinical AI assistant platform designed by and for healthcare professionals
Quadrivia builds Qu, a multi-modal clinical assistant with capabilities spanning diagnosis support, patient interactions, administrative tasks, and chronic care monitoring. The tech stack reflects a real-time-first architecture: Python + FastAPI + WebRTC + LiveKit for synchronous voice and chat, DSPy + RAG for reasoning, and event-driven data pipelines (Pub/Sub, Kubernetes) for agent orchestration at scale. The team is senior-heavy (8 of 9 known hires) and engineering-focused (7 roles), suggesting deep domain complexity in clinical AI safety and multi-agent coordination rather than a product-market-fit phase push.
Quadrivia develops Qu, a clinical decision-support and administrative assistant designed in collaboration with healthcare professionals. The platform spans patient-facing interactions (real-time chat and voice via WebRTC and SIP) and back-office clinical workflows (documentation, monitoring, care coordination). Core to the product is a multi-agent LLM architecture with RAG components for evidence-based reasoning and event-driven pipelines for throughput and reliability. The company operates from London with a small, senior-dominated engineering team and limited hiring velocity, indicating a focus on building foundational technology rather than rapid customer acquisition.
Python, FastAPI, Vertex AI, GCP services (Cloud Run, Pub/Sub, GCS), Kubernetes, Docker, LiveKit, WebRTC for real-time communication, DSPy and RAG for LLM reasoning, and React/TypeScript for frontend.
London, United Kingdom. The company also hires in Bulgaria and Spain.