Insights
Product, design, AI, and engineering perspectives from our team.
Technical Debt and AI: Why You Can't Build on a Broken Foundation
Layering AI onto systems carrying significant technical debt doesn't hide the debt — it amplifies it. This post covers how to audit the specific foundations your AI use case depends on, prioritise remediation without a full rewrite, and sequence modernisation and AI adoption so your investment actually reaches production.
AI Product Strategy: A Practical Guide for Australian Companies
A practical guide to AI product strategy for Australian mid-market companies — covering frameworks for identifying AI value, structuring a strategy engagement, common failure modes, and what to look for when selecting an external partner.
How to Evaluate an AI Development Company in Australia
Choosing the right AI development company in Australia is consequential. This guide gives mid-market technology leaders a practical checklist covering production references, demo-only portfolios, vendor lock-in, MLOps planning, and IP ownership — the areas where AI engagements most commonly go wrong.
The AI Hype Cycle: What's Real, What's Not, and What Matters
AI capabilities are real, but the gap between a compelling demo and a production system is wider than most vendors admit. This article maps which AI capabilities are genuinely production-ready, which are still maturing, and how to build an investment framework grounded in reality rather than hype.
Building AI Products on Supabase, PostgreSQL and pgvector
For lean engineering teams building AI features, Supabase and PostgreSQL have become a serious default stack. This article covers why the combination works, where pgvector fits into RAG architectures, how row-level security keeps AI features production-safe, and what the trade-offs look like as you scale.
Snowflake vs BigQuery for Australian Mid-Market Data Platforms
Snowflake and BigQuery are the two dominant cloud data warehouses for mid-market teams in Australia, but the right choice depends on your cloud estate, data-residency obligations, cost profile, and AI roadmap. This guide cuts through the noise to help Australian data and engineering leaders make a defensible decision.
MLflow vs Weights & Biases: Experiment Tracking and Model Registry
MLflow and Weights & Biases are the two platforms most growing ML teams evaluate for experiment tracking and model registry. This guide compares them honestly across deployment model, data residency, collaboration, and production reproducibility — so you can make the right call for your team and regulatory context.
PyTorch vs TensorFlow in 2025: Choosing a Production ML Framework
PyTorch and TensorFlow are both production-capable ML frameworks in 2025, but they suit different teams, workloads, and deployment environments. This guide helps technical leaders make a defensible framework choice based on ecosystem fit, serving requirements, and team context — not benchmarks or hype.
LLM Evaluation in Production: A Three-Layer Approach
Shipping an LLM feature is the easy part. Knowing whether it still works correctly six weeks later — after a prompt change, a model version bump, or a shift in user behaviour — is where most teams struggle. This post covers a three-layer evaluation approach that gives engineering teams real confidence in production LLM systems.