Infrastructure insights
from the field.

Connecting AI to your data is just another integration project

An AI system has no knowledge of your organisation until it's connected to your actual data. That connection follows the exact same architecture decisions — and the same scaling path from database to data warehouse to big data — that infrastructure teams already navigate.

Why AI decisions need the same governance as any change request

An AI system that recommends an outcome is not the same as one that decides an outcome. The governance discipline that keeps that distinction safe is the same discipline already applied to change management, access control, and budget management.

Why RAG performance depends on the same discipline as database indexing

AI retrieval systems slow down for the same reason poorly designed databases slow down — no indexing, no partitioning strategy. The fix is the same discipline infrastructure teams already apply to every database they operate.

Why AI search needs a different kind of database

AI systems don't search the way traditional databases search. Understanding why requires no data science background — just the same architectural thinking infrastructure teams already apply when choosing the right storage for the right workload.

The one design pattern behind every AI system

Every AI system relies on one architectural pattern infrastructure teams already use every day. Recognising it is what separates teams that can evaluate and operate AI from those dependent on a vendor.

How AI agents think — decision logic for infrastructure engineers

AI agents are not magic. They observe, plan, act, and reflect — the same structural logic infrastructure engineers have used for years. Here is how to think about agent reasoning without the hype.

Why infrastructure architects belong at the center of the AI deployment conversation

Generative AI and agentic AI get discussed almost entirely in data science and software terms. In regulated enterprise environments, the harder questions are architectural — and that's where this series starts.

Running Red Hat OpenShift on Nutanix — what actually changes

OpenShift on Nutanix AHV removes a lot of the operational complexity you expect from Kubernetes in production. Here's what the architecture looks like and where the edge cases hide.

Migrating to Microsoft Entra ID — the parts nobody warns you about

Entra migrations look clean on paper. In practice there are licensing traps, legacy authentication dependencies, and conditional access conflicts that surface at the worst moments.

Designing a DR strategy that actually meets your RPO

Most DR designs fail not during disasters but during the design phase — when RPO targets are set without understanding what replication technology can realistically deliver.

vSAN vs Nutanix — choosing the right HCI for your environment

Both are mature, proven platforms. The decision comes down to your existing stack, team skills, and where you expect to be in three years — not marketing benchmarks.

Why immutable backup storage should be non-negotiable in 2025

Ransomware doesn't just encrypt your data — it targets your backups first. Immutable storage changes the equation entirely, and the architecture is simpler than most teams think.

Scaling VMware Horizon without destroying your storage budget

Large VDI deployments punish under-engineered storage designs quickly. Here's how to size correctly from the start and where the common mistakes happen at 500+ seats.