Operations | Monitoring | ITSM | DevOps | Cloud

Traditional monitoring watches four things: whether the system is up (availability), how fast it responds (latency), how often it errors (error rate), and how hard the underlying resources are working (resource utilization). This system works because software is deterministic: the same input produces the same output, so a deviation means something broke, and it announces itself as a timeout.

Top Tips is a weekly column where we highlight what's trending in the tech world and share practical ways to stay ahead. This week, let's look at what happens when we fill every spare moment with digital content, why a little boredom might be useful, and how to make room for it in our screen-filled lives. Imagine you're waiting for your coffee. It's going to take three minutes, so you reach for your phone.

Remote work has changed more than where we work. It has reshaped how teams collaborate, how organizations manage performance, and how employees balance professional and personal responsibilities. Now, AI is adding another dimension to this transformation. From AI assistants that summarize meetings to automation tools that handle repetitive tasks, technology is making remote work more efficient in several ways.

Top tips is a weekly column where we highlight what’s trending in the tech world and share ways to stay ahead. This week, we’re exploring ways to consume information mindfully and avoid overstimulation. Today's internet user consumes content all the time. They listen to podcasts when they commute, watch YouTube when they do household chores, and listen to music when they work. The amount of content they consume is enormous, and the amount of content available is endless.

Most downtime doesn't come from problems your team can't solve. It comes from problems they've already solved sitting in a runbook or a text file, waiting for someone to run the fix manually. The fix is documented, and the steps are proven, yet the recovery still takes 40 minutes because a human has to log in and type it out. Workflow automation exists specifically to close that gap: The fix runs the moment the alert is triggered, not when someone sees it.

The moment your page slows down, two clocks start. One is yours: time to alert, time to investigate, time to fix. The other belongs to the user staring at the slow page. Yours is measured in minutes. Theirs runs out in seconds. That gap is what AI agents close. Zia Agents in OpManager Nexus detects an issue, works out the cause, and runs the fix on its own, often before your users feel a thing. This blog looks at how that changes the experience you deliver.

Top tips is a weekly column where we highlight what’s trending in the tech world and share ways to stay ahead. This week, we’re looking at what you can when you’re having a slow-paced, less hectic workweek. We’ve all been there at intermittent intervals of our jobs: You check into the office right after the weekend only to realize you’re having one of those slow, relatively low-pressure weeks. So what do you do to stay productive?

Generative AI (GenAI) is moving quickly into banking. According to a 2025 survey by McKinsey, 52% of financial institutions surveyed already consider GenAI a priority, while another 39% are interested but have not yet made it a top priority. As adoption grows, banks need to think carefully about what AI agents can access, which identities it uses, what actions it can take, and whether those activities can be traced when something goes wrong.

Azure Monitor doesn't have a price. It has a bill and those are very different. There's no plan tier to pick, monthly set fee, or simple number to sanity-check against your budget. Instead, you're charged across a handful of separate meters: log ingestion, log queries, retention, alert rules, web tests, and custom metrics. Each of these metrics scale independently. Most teams don't see the full cost until the bill arrives.

You bought the GPUs to make AI training faster. So why are they sitting idle? When GPU utilization drops during a training run, the obvious answer is to blame the accelerator. Maybe the workload is too small. Maybe the GPU isn't powerful enough. Maybe it's time to add more hardware. But what if the GPU isn't the problem at all? A training workload is only as fast as the infrastructure feeding it.