Skip to main content
AI at Oxford

Streamlining IT troubleshooting and building new skills

User Case Study

IT Support Staff

Before using GenAI tools, my process for troubleshooting IT issues would be to search online using Google, often ending up on a trail of Reddit threads or Microsoft forum pages. Many of these contained the same questions but no clear answers, or replies that were vague or condescending. Alternatively, I could ask a colleague and wait, if they were available. None of this was particularly efficient.

GenAI has changed that. When I ask an AI tool an IT question, it usually gives me a direct and usable answer. If I reply with, “I have tried that”, it will suggest further troubleshooting steps. This back and forth is far quicker than traditional searching and feels more like a conversation with a knowledgeable colleague who actually responds.

Alongside troubleshooting, I've been using GenAI to develop my IT skills. For example, I’ve been working on a personal project to build a network monitor that alerts me via Alexa and Telegram if a server goes down. I use  ChatGPT to produce and explain code, which has been especially helpful given that I have very limited coding experience. This has turned into a larger project than expected, and I'm currently taking a break from it.

I also regularly use Copilot to troubleshoot issues. Because it's tied into Nexus365, it's particularly useful for tech support queries. ChatGPT, however, is better for coding and for helping me create small projects from scratch when I would not otherwise know where to begin.

Since adopting GenAI, I've seen clear benefits. As someone working in tech support, if I receive a question that I'm unsure about, I can ask AI to explain the issue or help me construct an answer. It's also been helpful for writing documentation, especially when an issue is complicated and the solution needs to be shared with others. Overall, using AI rather than a normal search engine has saved a significant amount of time that would otherwise be spent sifting through largely irrelevant or low-quality results.