Technology briefing

Technology, cloud, infrastructure and practical geekery

A practical map of enterprise technology, architecture and the decisions that matter after deployment.

MicrosoftCloudInfrastructureHow-ToSecurityJersey technology

Technology is easiest to understand when it is treated as a connected system rather than a stream of product announcements. Cloud changes the network. Identity changes the security model. Virtualisation changes storage and recovery. AI changes the way information is found and used. A licensing change can alter an architecture decision just as quickly as a new technical capability.

This is the top-level map for technology on isageek. It brings together the subjects I write about most — Microsoft, cloud, enterprise infrastructure, security, digital platforms and the more practical side of geekery — and focuses on the decisions underneath them: what a technology is actually useful for, what it depends on, what can go wrong, and how I would think about operating it after deployment.

Last reviewed: 5 September 2026. Supporting guides are refreshed as the site grows.

The technology landscape I care about

Technology is an architecture problem before it is a product problem

A lot of technology coverage starts with a feature list. That is useful up to a point, but most expensive mistakes happen between products rather than inside them. A new platform may be technically excellent but create an awkward identity dependency. A cloud service may remove server maintenance while increasing the importance of connectivity and access control. A storage refresh may look like a capacity project until recovery objectives expose the real requirement.

The useful questions therefore tend to be architectural: what problem are we solving, what dependencies are we creating, how will this be secured, how will it be operated, and how will it recover? Those questions apply whether the technology is a Microsoft service, a hypervisor, an AI platform, a backup repository or a home-lab experiment.

Microsoft: the ecosystem effect

Microsoft technology deserves a large part of this site because its products increasingly behave as one connected operating environment. Entra ID affects Microsoft 365 and Azure. Conditional Access affects how users reach applications. Defender increasingly spans endpoints, identity, cloud and email. Azure can host infrastructure, but it can also provide monitoring, security, identity and recovery for systems that never move out of a local datacentre.

That means a Microsoft decision should rarely be assessed in isolation. The real value — and sometimes the real complexity — appears in the relationships between identity, endpoint management, collaboration, cloud services and traditional infrastructure.

Cloud: placement, not ideology

“Cloud first” and “keep it on-premises” are both poor substitutes for architecture. Some workloads are excellent SaaS or public-cloud candidates. Others remain easier to operate, cheaper or more resilient in local infrastructure. Many established organisations inevitably become hybrid because their application estate, data, risk profile and replacement cycles do not all move at the same speed.

I therefore treat cloud as a workload-placement and operating-model question. Latency, data handling, identity, egress, recovery, supportability, licensing and staff capability matter just as much as compute pricing. Good cloud design is not measured by the percentage of servers moved off-site; it is measured by whether the resulting service is simpler, safer, supportable and recoverable.

Infrastructure still matters

Infrastructure periodically gets declared obsolete. It never quite cooperates. Virtual machines still need compute. Data still needs storage. Applications still need networks. SaaS platforms still depend on identity and connectivity. Recovery still requires somewhere trustworthy to restore from. Even serverless services sit on somebody else’s infrastructure.

The interesting change is that infrastructure has become broader. It now includes physical platforms, software-defined storage, cloud networking, identity dependencies, immutable backup, automation, observability and the operational controls around them. The goal is not to preserve old infrastructure for its own sake; it is to understand the foundation beneath the service.

Digital, AI and emerging technology

“Digital” can become meaningless if it is used as a synonym for anything involving a computer. I use it more narrowly: the way technology changes a service, organisation or economy. That includes digital identity, automation, online services, data platforms and the infrastructure that allows them to operate reliably.

AI belongs in that discussion as a capability rather than a novelty. The practical questions are increasingly about trusted data, permissions, governance, model behaviour, workflow integration and whether an AI system produces measurable value. The technology is moving quickly, but the disciplines around deployment — security, evidence, recoverability, cost and human accountability — remain familiar.

How I evaluate a technology change

When a product or platform looks interesting, I find it useful to strip away the launch language and work through a consistent set of questions:

1. What problem disappears? If the answer is vague, the change is probably being led by the product rather than the requirement.
2. What new dependency appears? Every simplification moves complexity somewhere else.
3. What happens on the bad day? Authentication failure, provider outage, ransomware, configuration error and human mistakes belong in the design.
4. Who operates it at 02:00? A platform that only one specialist understands carries a different risk from one the wider team can support.
5. How do we leave? Data portability, standards, contractual terms and replacement cost matter before a platform becomes critical.
6. What evidence tells us it worked? Performance, reliability, cost, user experience or risk should improve in a way that can be measured.

Practical guides and geekery

Not everything needs to be enterprise architecture. Building a home server, comparing a gadget, experimenting with Proxmox or working out why something refuses to connect can teach the same habits on a smaller scale: understand the components, change one thing at a time, measure the result and keep a route back.

That is why “geekery” stays in the mix. It is part of the site’s character, but the aim is still useful technical content rather than novelty for its own sake.

Where to start

If you are here for enterprise technology, start with the three specialist pillars above. If you are interested in the Channel Islands, Jersey’s Digital Edge tracks how these same cloud, infrastructure, security and AI questions play out in a small but unusually technology-dependent jurisdiction.

The individual articles go deeper; this page exists to keep the map coherent.

About this guide

isageek is an independent technology site focused on the practical side of modern IT: architecture, infrastructure, cloud, Microsoft platforms, resilience and the occasional diversion into broader geekery. The emphasis is on consequences and trade-offs rather than reproducing product announcements.