Andrew Stewart

About

Google Health engineer who owned and shipped an AI-powered clinical insights dashboard and NLP clinical note extraction service end-to-end (React/Next.js frontend; Python/Node microservices on GKE; TensorFlow transformers; BigQuery analytics). Demonstrated strong production rigor (CI/CD, testing, observability, guardrails for sensitive data) and delivered measurable outcomes including 30% faster diagnostics, 40% less manual documentation, 15% higher adoption, and 25% lower ops costs.

Key Strengths

• Owned end-to-end production feature across frontend/backend/data/infra (React/Next.js + Python/Node microservices on GKE)
• Built real-time clinician dashboard with GraphQL subscriptions and data visualizations (D3, Material UI)
• Productionized NLP clinical entity extraction (diagnoses/medications/symptoms) using TensorFlow transformer models with standardized JSON outputs
• Established rigorous evaluation and monitoring (precision/recall benchmarks on annotated notes; Prometheus/Grafana/Stackdriver observability)
• Implemented safety/guardrails for sensitive healthcare data (access controls, audit logging, confidence thresholds with clinician review)
• Modernized legacy monolith to microservices via phased migration and dual-running with regression testing and feature toggles
• Demonstrated measurable impact: 30% reduction in diagnostic time; 15% increase in platform adoption; 40% reduction in manual documentation time; 25% operational cost reduction
• Partnered with CTO/CPO and clinical stakeholders to translate ambiguous goals into roadmap and milestones; balanced AI accuracy vs delivery speed
• Effective cross-team execution with stable API contracts, clear interfaces, and structured coordination (weekly syncs, Jira)

Experience

Education

Skills

Languages