Field notes · Mackay, QLD
Insights & innovation.
Practical writing on local AI, open-source models and the unglamorous work of making technology run in a real business. No hype cycle required.
Latest posts
Latest posts
Why I Wrote Be Practical
A small-town founder story about owning AI, surviving a tough diagnosis, and turning AI into a reliable operating advantage.
Engineering Local AI: Faster Proposals, Cleaner Documentation, Less Rework
Turn PDFs, standards and project history into a private knowledge assistant with citations.
Hospitality & Tourism Local AI: Guest Experience Without the Data Leakage
Guest messaging, virtual concierge and feedback curation that keeps guest data inside your environment.
Manufacturing Local AI: Reduce Downtime, Capture Knowledge, Standardize Quality
Technician copilots grounded in SOPs, predictive maintenance insights and quality standardisation.
Small Business Local AI Setup: The Simplest Stack That Actually Works
One machine, one interface, and workflows that save hours every week with Ollama and Open WebUI.
Enterprise AI Infrastructure: Private LLMs That Don't Break in Production
An on-prem reference architecture: inference engines, routing, RBAC, auditability and observability.
Local AI Computers: What to Buy (and Why Specs Actually Matter)
VRAM, bandwidth, fast NVMe and building a machine where your models fit and responses stay fast.
Open-Source Alternatives to Paid Apps: The Practical Stack
Replace expensive SaaS by focusing on data ownership, exportability and operational simplicity.
Open-Source LLMs: What to Run Locally (and What to Serve in Production)
How to choose the right local and production inference stack — Ollama, vLLM and llama.cpp.
Open WebUI: Your Self-Hosted ChatGPT Interface for Local & Private LLMs
A practical way to give your team a modern chat interface while keeping your stack flexible.
Need it built?
Reading is good. Running is better.
We turn the theory into a working system on your own hardware — documented, owned and supported.