Skip to content
Ryan James
Menu

← Newsletter2 min read

Transforming IT Support with AI: How I Used Google NotebookLM to Build an Intelligent Help Desk

As an IT Administrator, my role is a constant balancing act between maintaining complex infrastructure and empowering 150 employees to use our technology effectively. Between managing documentation and leading training sessions, the volume of “how-to” knowledge I’m responsible for can feel endless.

Traditional AI help desks often fall short in this kind of environment. If you’ve interacted with one in a technical setting, you know the frustration: a generic article gets surfaced, it doesn’t solve the problem, and suddenly you’re 15 minutes into an automated loop before you can reach a human. I knew there had to be a better way to give colleagues immediate, accurate answers — without adding more to my team’s workload. That’s when I turned to Google NotebookLM.

The Challenge: Knowledge at Scale

Like many IT professionals, I sit on a mountain of content:
∙ In-depth technical articles
∙ Step-by-step how-to guides
∙ Standard Operating Procedures (SOPs)
∙ Training slide decks
∙ External support documentation for our tech stack, including Google Workspace

The problem was never a lack of information — it was accessibility. People often forget how to perform infrequent tasks (like retrieving transcribed voicemails) and need a quick answer without diving into a support portal rabbit hole.

The Solution: A Custom-Trained AI Notebook

Using NotebookLM, I consolidated every resource I’d created into a single, intelligent notebook. I then shared it with staff through a link in my email signature and on our IT support site — no login required, no friction.

The results have been genuinely transformative:
∙ Context-specific answers: Employees get responses grounded exclusively in the documentation I’ve provided — no hallucinated advice, no irrelevant results.
∙ Fewer support tickets: Common “low-hanging fruit” questions are answered instantly, freeing the IT team to focus on higher-complexity work.
∙ Smoother onboarding: New hires now have a comprehensive, searchable knowledge base to help them answer employee questions from day one.
∙ Zero-maintenance updates: When I update an article or SOP, the AI reflects those changes automatically — no retraining required.

Why This Matters for the Future of IT

We’re entering an era of agentic AI, where tools like Gemini, Claude, and ChatGPT are growing increasingly capable. But for IT administrators, the real advantage lies in guardrails — the ability to ensure an AI only surfaces answers from your verified content sources.

Google NotebookLM is ahead of the curve here. It offers a more powerful and approachable way to leverage large language models (LLMs) for specific organizational needs than most “custom GPT” or skills-based tools I’ve evaluated.
Whether you’re in education or the corporate world, I’d strongly encourage you to explore how these tools can transform your sprawling documentation into a conversational, 24/7 resource for your team. The technology is ready — the only question is whether your knowledge base is.

Keep coding!
Ryan

This post first went out on Substack. Join the conversation there or subscribe to get the next one by email.