
Software, AI and IT insights
Writing from PicNet's engineers on AI, software, cloud and the systems Australian organisations run on.
Scanning public health signals: outbreak and drug safety news monitoring
How to build an AI monitor for WHO, TGA and state health alerts: feed ingestion, tiered classification, routing to the right team, and honest precision and recall trade-offs.
#public-health-surveillance#drug-safety#alerting#llm-classification#integration
A policy and procedure chat assistant for health organisations
How to build a grounded RAG assistant over internal health policies: citations, refusals, why RAG beats fine-tuning, and the evaluation you run before rollout.
Where deterministic matching ends: cleaning and matching patient records with AI
How to resolve duplicate patient and customer records across systems: deterministic rules first, LLM adjudication on the grey zone, human sign-off.
#entity-resolution#data-integration#centazio#master-data#fhir
From meeting audio to structured minutes in health settings
How to turn committee and case conference audio into structured minutes in Australian health services: consent, accuracy review, retention and real error rates.
#meeting-transcription#speech-to-text#clinical-governance#privacy#ai-governance
Mining patient feedback and complaints for themes
How Australian health services can turn years of free-text patient feedback and complaints into themes and risk signals, with an architecture, costs and limits.
#patient-feedback#natural-language-processing#clinical-governance#privacy#large-language-models
AI-assisted medical coding: suggestions your coders can defend
How to build an AI assistant that suggests ICD-10-AM and AR-DRG codes for coder review in Australian hospitals: architecture, audit trails, costs and honest limits.
#clinical-coding#icd-10-am#activity-based-funding#human-in-the-loop#health-it