
Software, AI and IT insights
Writing from PicNet's engineers on AI, software, cloud and the systems Australian organisations run on.
Your AI is only as good as your data: why integration comes first
Most failed AI pilots are data failures, not model failures. What AI-ready really means, the integration architecture behind it, and a checklist before you fund.
#data-integration#master-data#ai-readiness#centazio#data-quality
Coding agents in a production .NET shop: what holds up
How PicNet runs Claude-style coding agents on production .NET code: standards docs in the repo, tests as the safety net, AI review first, and what agents still get wrong.
#coding-agents#dotnet#code-review#developer-productivity#ai-engineering
One small team, 100+ automated checks: anatomy of an AI-assisted monitoring platform
How PicNet's in-house monitoring platform runs 100+ scheduled checks with OK/ALERT/FAULT states, LLM-written summary emails, a watched heartbeat, and what it costs.
Practical AI in DevOps: the series
A series on how AI agents and automation change day-to-day software operations, drawn from PicNet's own stack: working systems, real costs, real limits.
Listening at scale: monitoring industry forums and communities with AI
How to monitor industry forums with AI classification instead of keyword alerts, encode per-community etiquette, and keep a human writing every reply.
#social-listening#community-marketing#brand-monitoring#llm-classification#content-strategy
Answering from your own documents: a policies and procedures assistant for any business
How to build a retrieval assistant that answers staff questions from your own HR, ops and product documents, with citations, refusals, evaluation and costs.