Insights

The thinking
behind the work

Deep dives on AI infrastructure, agentic systems, and data architecture — written from the field, not the whiteboard.

Beyond the Chain: Engineering Production-Grade Multi-Agent Systems
Agentic Systems
01
19 June 20267 min read

Beyond the Chain: Engineering Production-Grade Multi-Agent Systems

The industry's focus is shifting from single-agent performance to multi-agent system reliability. We dissect the critical engineering patterns for orchestration, evaluation, and data feedback that separate production-grade systems from brittle prototypes.

Read full article
The Industrialisation of AI Agents: Navigating the 2026 Platform Shift
AI Platform Strategy
02
18 June 20266 min read

The Industrialisation of AI Agents: Navigating the 2026 Platform Shift

The era of agentic AI prototypes is over. Recent updates from Azure, Google, and the automation ecosystem signal a shift to industrial-scale agent platforms. This analysis separates the durable architectural trends from the hype, guiding CTOs on what to build next.

Read article
The BI Re-architecture: AI is Forcing a Reckoning for Semantic Layers and Analytics Workflows
BI Strategy
03
18 June 20268 min read

The BI Re-architecture: AI is Forcing a Reckoning for Semantic Layers and Analytics Workflows

The latest wave of generative AI embedded in BI tools is not an incremental update; it's a fundamental architectural shift. For technical leaders, the focus must move from dashboards to the semantic core that governs this new agentic layer.

Read article
The End of the Dashboard: Navigating the Generative BI Paradigm Shift
BI Strategy
04
18 June 20267 min read

The End of the Dashboard: Navigating the Generative BI Paradigm Shift

Recent updates to Power BI, Looker, and QuickSight signal a fundamental shift from interactive to generative BI. For analytics leaders, the era of dialogue-driven analysis is here, and the semantic layer is now the most critical component of your data stack.

Read article
The Two-Speed Data Architecture: Unifying BI and AI in the Consolidated Lakehouse
Data Architecture
05
18 June 20267 min read

The Two-Speed Data Architecture: Unifying BI and AI in the Consolidated Lakehouse

The industry's rush to consolidate data platforms into a single lakehouse for both BI and AI overlooks a fundamental architectural conflict. The slow, governed world of analytics and the fast, iterative demands of AI require a deliberate two-speed architecture to coexist without compromising performance or governance.

Read article
Beyond the Monolithic Lakehouse: Architecting the Polyglot Serving Layer for AI
Data Architecture
06
18 June 20269 min read

Beyond the Monolithic Lakehouse: Architecting the Polyglot Serving Layer for AI

The lakehouse is necessary but not sufficient for the demands of modern AI. We break down the architectural imperative for a specialised, polyglot serving layer—combining real-time OLAP, vector search, and direct lakehouse access—unified by a robust semantic layer.

Read article
The Post-Unification Stack: Architecting Data Platforms for the Agentic Era
Data Architecture
07
17 June 202611 min read

The Post-Unification Stack: Architecting Data Platforms for the Agentic Era

The traditional divide between transactional and analytical systems is collapsing under the demands of AI agents. We dissect the new "Lake Transactional/Analytical Processing" (LTAP) paradigm and outline the architectural blueprint for building a data platform that can power AI that doesn't just analyse, but acts.

Read article
Beyond the AI Factory: Architecting for Sustained Inference Throughput
AI Infrastructure
08
15 June 20269 min read

Beyond the AI Factory: Architecting for Sustained Inference Throughput

The era of the 'AI Factory' is here, but capital investment in GPUs is only the first step. The critical challenge for senior technologists is architecting the software stack to prevent these billion-dollar assets from becoming monuments to inefficiency. We dissect the critical trade-offs in quantisation, serving engines, and advanced throughput techniques that define modern AI infrastructure.

Read article
The Observability Crisis in Agentic AI: From Output Metrics to Behavioural Tracing
LLM Evaluation
09
12 June 20267 min read

The Observability Crisis in Agentic AI: From Output Metrics to Behavioural Tracing

The industry's rush to deploy agentic AI has created a critical observability gap. Legacy evaluation metrics designed for simple RAG are failing to capture the complex failure modes of multi-agent systems. This article deconstructs the shift from output-focused evaluation to process-centric observability, outlining the engineering patterns required to build, debug, and productionise reliable agents.

Read article
The Semantic Lakehouse: Architecting the Data Foundation for an Agentic AI Future
Data Architecture
010
10 June 20268 min read

The Semantic Lakehouse: Architecting the Data Foundation for an Agentic AI Future

The era of autonomous AI agents demands a fundamental architectural shift beyond the traditional lakehouse. We dissect the Semantic Lakehouse—a new pattern unifying open table formats with a machine-readable semantic layer—as the essential foundation for reliable, governable, and effective agentic AI.

Read article
The AI Factory's Bottleneck: Architecting the Inference Stack for Compound Agentic Latency
AI Infrastructure
011
8 June 20268 min read

The AI Factory's Bottleneck: Architecting the Inference Stack for Compound Agentic Latency

The era of measuring LLM performance by simple throughput is over. As enterprises build 'AI Factories' for agentic workloads, the critical bottleneck is now 'compound latency' — the end-to-end time for complex, multi-step tasks. This requires a fundamental rethink of the inference stack.

Read article
Beyond the Prototype: Engineering Reliability in Agentic AI Systems
LLM Engineering
012
5 June 20267 min read

Beyond the Prototype: Engineering Reliability in Agentic AI Systems

The leap from a compelling agentic AI demo to a production-grade system is a chasm of engineering challenges. We dissect the three pillars of reliability that separate robust enterprise agents from brittle prototypes: rigorous evaluation, targeted alignment, and resilient tool orchestration.

Read article
Beyond the Ecosystem: Architecting the Open Data Foundation for Enterprise AI
Data Architecture
013
3 June 20267 min read

Beyond the Ecosystem: Architecting the Open Data Foundation for Enterprise AI

The industry is rushing to adopt vendor-led AI ecosystems, spurred by announcements at Snowflake's Data Cloud Summit. This article argues for a different approach: building a sustainable AI-native platform on an open, interoperable foundation using technologies like Apache Iceberg and a decoupled semantic layer to avoid lock-in and ensure long-term architectural integrity.

Read article
Deconstructing the Agentic Stack: Inference Architectures for Compound AI
AI Infrastructure
014
2 June 20268 min read

Deconstructing the Agentic Stack: Inference Architectures for Compound AI

The era of agentic AI demands we move beyond optimising single, monolithic models. We must now architect for 'compound AI'—complex graphs of heterogeneous models and tools. This requires a fundamental rethink of our inference stacks, from GPU scheduling to cost attribution.

Read article
The Compute Locus Problem: Architecting for Hybrid AI Inference
AI Infrastructure
015
2 June 20268 min read

The Compute Locus Problem: Architecting for Hybrid AI Inference

The era of monolithic AI endpoints is over. Spurred by enterprise needs for data sovereignty and cost control, the new frontier is hybrid inference orchestration. We dissect the architectural patterns required to build systems that dynamically choose where to run AI workloads—on-device, on-prem, or in the cloud.

Read article
The Great Bifurcation: Architecting for Centralised vs. Decentralised AI
AI Infrastructure
016
2 June 20268 min read

The Great Bifurcation: Architecting for Centralised vs. Decentralised AI

The era of cloud-only AI inference is over. Powerful deskside hardware forces a critical architectural decision: centralise for throughput or decentralise for latency? We dissect the trade-offs that senior technical leaders must now navigate.

Read article
The Rubin Effect: Rethinking Your AI Platform for the Agentic Era
AI Infrastructure
017
2 June 20266 min read

The Rubin Effect: Rethinking Your AI Platform for the Agentic Era

NVIDIA's Vera Rubin platform signals a paradigm shift from simple LLM inference to complex agentic AI workloads. This demands a fundamental rethink of your infrastructure, moving beyond stateless endpoints to stateful orchestration engines that can manage long-running, multi-step tasks.

Read article
NFP Donor Retention Analytics: Moving Beyond Spreadsheets
NFP Analytics
018
26 Mar 20267 min read

NFP Donor Retention Analytics: Moving Beyond Spreadsheets

Australian NFPs lose more than half their donors every year — not because of poor fundraising, but because the data to predict and prevent churn sits in disconnected silos. Here is how to build the analytics foundation that turns retention from a guess into a strategy.

Read article
Automated FDS Compliance Dashboards for AFSL Holders
Financial Data
019
25 Mar 20267 min read

Automated FDS Compliance Dashboards for AFSL Holders

Manual Fee Disclosure Statement generation costs Australian financial planning practices between 400 and 800 hours per year — and creates the compliance gaps ASIC is actively looking for. Here is the architecture that eliminates both the cost and the risk.

Read article
Automating ACARA Reporting for Independent Schools
Education Data
020
24 Mar 20266 min read

Automating ACARA Reporting for Independent Schools

Independent schools spend an average of 43 staff hours per term on government reporting — data that already exists in their systems, locked behind eight disconnected platforms. Automation is not a future aspiration; it is a practical project with measurable ROI from the first submission cycle.

Read article
Xplan & Midwinter Data Integration with Power BI
Data Integration
021
23 Mar 20266 min read

Xplan & Midwinter Data Integration with Power BI

Xplan manages your clients. Midwinter models their futures. Power BI answers the question neither platform was built to ask: how is your practice actually performing? This is the integration architecture that connects all three — and the dashboards that make the investment visible from day one.

Read article
Real-Time AI: Building Streaming Pipelines That Actually Feed Your Language Models
Data Engineering
022
22 Mar 20268 min read

Real-Time AI: Building Streaming Pipelines That Actually Feed Your Language Models

Most enterprise AI systems are running on stale data — not because teams didn't think about freshness, but because the default architecture is batch-oriented. Real-time AI closes the gap between what your models know and what is actually happening in your systems right now.

Read article
Gemini 3.1, GPT-5.4, and Claude Opus 4.6: What the New Frontier Means for Enterprise AI
LLM Models
023
13 Mar 20268 min read

Gemini 3.1, GPT-5.4, and Claude Opus 4.6: What the New Frontier Means for Enterprise AI

Three frontier models, three different bets on what enterprise AI needs most. Gemini 3.1 pushes native multimodality, GPT-5.4 targets agentic reliability, and Claude Opus 4.6 leads on deep reasoning and safety. Here is what the new frontier means for your architecture.

Read article
MCP and A2A: The Protocols Shaping How AI Agents Communicate
Protocols
024
10 Mar 20267 min read

MCP and A2A: The Protocols Shaping How AI Agents Communicate

The AI agent ecosystem is fragmenting at exactly the wrong time. The Model Context Protocol and the Agent-to-Agent protocol are emerging as the infrastructure layer enterprise AI has been missing — standardising how models connect to tools and how agents coordinate with each other.

Read article
Retrieval-Augmented Generation in Production: Beyond the Proof of Concept
RAG
025
9 Mar 20267 min read

Retrieval-Augmented Generation in Production: Beyond the Proof of Concept

Most enterprise RAG systems underperform not because the architecture is flawed, but because the path from demo to production exposes a stack of decisions a prototype never surfaces — from chunking strategy and embedding choice to reranking and graceful failure handling.

Read article
Architecting Agentic Workflows for Enterprise Scale
AI Agents
026
28 Feb 20268 min read

Architecting Agentic Workflows for Enterprise Scale

Most enterprise AI pilots stall not because the models aren't capable — it's because the architecture around them can't hold the weight. Building agentic systems that survive contact with real-world complexity requires rethinking how tasks are decomposed, how agents communicate, and where humans remain in the loop.

Read article
Optimizing AI Infrastructure: GPU Clusters and Vector Databases
Infrastructure
027
14 Feb 20266 min read

Optimizing AI Infrastructure: GPU Clusters and Vector Databases

The gap between a demo that impresses and a system that performs at scale almost always comes down to infrastructure choices made too early. From GPU cluster topology to vector index sharding strategies, the decisions you make at the infrastructure layer set hard ceilings on everything above.

Read article
The Evolution of Data Architecture in the Age of LLMs
Architecture
028
31 Jan 20267 min read

The Evolution of Data Architecture in the Age of LLMs

The data stack that served us well for a decade is showing its age. LLMs don't just consume data differently — they demand a fundamentally different approach to how data is stored, enriched, retrieved, and served. The rise of semantic layers and RAG-first design is a structural shift.

Read article
From Prototype to Production: Deploying AI Systems at Scale
Engineering
029
17 Jan 20269 min read

From Prototype to Production: Deploying AI Systems at Scale

Getting an AI model to work in a notebook is easy. Getting it to work reliably, cost-effectively, and safely in production is a different discipline entirely. MLOps isn't DevOps with a model attached — it's a continuous negotiation between experimentation velocity and operational stability.

Read article
The Strategic Imperative of an AI-First Data Culture
Strategy
030
3 Jan 20265 min read

The Strategic Imperative of an AI-First Data Culture

Technology is rarely the binding constraint in an AI transformation — culture is. The organisations moving fastest aren't those with the most sophisticated models; they're the ones that have made data fluency a first-class organisational capability and built governance structures that let AI operate with confidence.

Read article