There’s a big problem, and its hiding in plain sight across the IT services industry. It isn’t a technology issue, but a knowledge issue. Critical operational insights, which keep services running smoothly, are often locked away within individual engineers, stove-piped tools and disconnected teams. As a consequence, these insights stay beyond reach for the professionals who need them most. Gartner has flagged this as a growing issue, reporting that nearly half of digital workers (47%) struggle to find the information they need to do their jobs effectively.
The consequences of these “knowledge silos” compound quickly. Efficiency and collaboration suffer, innovation stalls and risks start to appear. Making matters worse, organisations become overly reliant on the knowledge of a select few engineers. This dependency on a handful of engineers leads to serious vulnerability the moment one engineer leaves the team. As enterprise environments grow more complex and data volumes continue to accelerate, the single-human dependency model of knowledge simply doesn’t scale.
How knowledge silos take hold
Knowledge silos don’t appear overnight; they develop as technology moves faster than organisations can document and share what they know. As a byproduct of this, essential operational knowledge splinters, residing in personal notes, hard-drives and inside the minds of engineers.
When an experienced engineer leaves the team, they take irreplaceable context with them. Knowledge regarding how systems behave under pressure, how incidents actually unfold, and which fixes genuinely work versus the ones that just look good on paper. Very few of these types of insight get captured in a form the rest of the team can access and reuse, and this quietly erodes organisational performance over time. Research has shown that employees can lose almost 2 hours a day, that’s 52 days a year or 2 person months annually, searching for the right information. This wasted time leads to slower decisions and weaker business outcomes across the board, not to mention reducing the cost effectiveness of the whole team.
Augmented AI changes the equation
AI has the potential to dissolve these silos, but only if it’s used as a layer that augments human expertise, not one that tries to replace it. When applied correctly, augmented AI serves as a living intelligence layer, absorbing context and continuously learning from operational data and the day-to-day experience of engineers, capturing every intention, every action, every outcome rather than letting them disappear. The payoff is substantial:
- Faster, more accurate resolutions through contextual guidance grounded in historical incidents
- Higher performance from less experienced engineers, who can operate with AI‑driven insight.
- More time for senior engineers to focus on strategic, high‑value work.
- Efficiency gains of 60–65% across support operations
Retrieval-Augmented Generation (RAG) does much of the heavy lifting here. Where traditional enterprise search leans on keyword matching and frequently produces irrelevant or stale results, RAG works differently. It interprets what it’s being asked, pulls only the most relevant knowledge, and layers in context before generating a response. The answer given by the automated system is rooted in an organisation’s own operational history.
Technology alone cannot get you there
As AI adoption continues to accelerate, a disparity between how much organisations are investing in the technology and how capable they actually are to apply it effectively is appearing. Deploying AI without skilled engineers steering the process increases the chances of new risks being introduced. This includes downtime, potential for misconfigurations, security gaps, and workflows that quietly become less efficient rather than more.
The truth is that AI is only as strong as the data that it learns from, the workflows it’s built into, and the people who shape, validate and refine it. Organisations need more than just the technology itself. They need engineers who understand service desk dynamics, recognise incident patterns, and grasp the practical constraints of enterprise environments.
As IT environments grow more complex and the volume of data keeps rising, organisations that don’t address their knowledge silos will find themselves falling behind. The way forward isn’t choosing between human expertise and AI; it’s effectively combining both. Building systems that don’t just store knowledge but actively improve how it’s accessed, applied and refined over time. Augmented AI is what turns fragmented data into insights you can actually act on. The organisations that invest in both the technology and the engineering talent to effectively operationalise it, are the ones that will pull ahead.
