SENIOR ML ENGINEER · SYDNEY
I’m Luke (Guanhua) Cao, a senior machine learning engineer. I build production ML systems that turn complex health data into better decisions.
Thoughtful engineering.
Meaningful clinical impact.
01 / SELECTED WORK
From anticipating patient risk to making sense of clinical records. A selection of my work in healthcare AI.
RECENT WORK / 2025
AI-powered analytics for better hospital performance. A collaboration between Beamtree and the NHS Confederation, bringing benchmarking and predictive forecasting to NHS trusts.
Meet the Evolve Collaborative (opens in a new tab)
PREDICTIVE ANALYTICS
Machine learning for the Health Roundtable Insights Platform, helping turn hospital performance data into useful forecasts.
Help healthcare teams anticipate changes in performance, clinical quality, patient safety, and operational efficiency.
Developed classical machine learning models to power predictive forecasting within an interactive hospital analytics platform.
CLINICAL NLP
Extracting medical information from unstructured notes to support clinical review and better connected patient data.
Important patient information lives in free-text notes that are difficult to analyse consistently at scale.
Clinical NLP to convert unstructured text into structured information and surface potential medication interactions and adverse events for review.
PATIENT MONITORING
Analysing vital signs and laboratory results to identify deterioration risk and support timely clinical intervention.
Give clinical teams an earlier, actionable view of patient risk across a ward or hospital.
Real-time patient monitoring that brings vital signs and laboratory results together to support risk assessment.
Product evidence: Beamtree reports a 5% reduction in length of stay in a published ADI implementation study. This is a product-level result, not an individual performance claim.
GENERATIVE AI
Combining LLM agents with retrieval-augmented generation to turn clinical records into ICD and procedure codes.
Clinical coding is a detail-intensive process that depends on understanding both the record and the relevant coding guidance.
Applying language models and retrieval to support code generation, reduce repetitive coding work, and keep relevant clinical context close to the task.
Project visuals are conceptual illustrations. Follow the links for product information.
02 / EXPERTISE
My work connects machine learning methods with the information and decisions that matter in healthcare.
Bringing machine learning into practical applications, with the operational context to make it useful.
Finding patterns in complex data and building models that help teams plan ahead.
Making clinical text more useful, from extracting information to automating coding workflows.
Working with clinical data to support interpretable, practical decisions at the point of care.
03 / A LITTLE ABOUT ME
Luke Cao.
Guanhua Cao / SydneyI’m a senior machine learning engineer at Beamtree in Sydney, focused on a question that keeps my work grounded: how can this help the people delivering care?
My experience spans hospital forecasting, clinical text analysis, predictive monitoring, and automated clinical coding. Across these areas, I focus on accurate, interpretable models and applications that fit the healthcare decisions they support.
I enjoy working at the meeting point of complex data and practical problems—turning the former into something people can actually use.
EXPERIENCE / BEAMTREE
My current work centres on production machine learning and the healthcare applications it supports—from hospital forecasting to clinical coding.
Full experience on LinkedIn (opens in a new tab)CURRENT EMPLOYMENT
Beamtree · Sydney, Australia
I work on production ML systems, MLOps, and applied AI for healthcare, connecting complex clinical and operational data with practical tools.
Predictive modelling for Health Roundtable, alongside the Evolve Collaborative’s focus on hospital performance and productivity.
Explore forecasting projectsClinical information extraction and automated coding, applying language models and retrieval to turn records into structured information.
Explore language AI projectsPredictive patient monitoring that brings vital signs and laboratory results together to help identify deterioration risk.
A cloud NLP service for creating, training, and managing machine learning models.
Machine learning to identify COVID-19 infection regions in lung CT images.
Education
04 / START A CONVERSATION
Working on a healthcare challenge, building an AI team, or exploring an idea? I’d love to hear about it.
[email protected]Based in Sydney, Australia.