Setting the Stage: The 20-Minute Board Challenge

At Bett Asia 2026 during the Google for Education Tech User Labs session, the room was packed with CTOs and university IT directors as presenters Ricky and Nanda kicked off the hands-on Tech User Lab. They set a realistic challenge for everyone in the room: imagine you are a university IT team tasked with analysing sprawling institutional data—from cloud clusters to student satisfaction surveys—and putting together a strategic proposal for the board under severe time pressure.

Ricky laid out a four-part roadmap for the session, focusing the hands-on segment on Targeted Investigation and Cross-Document Synthesis, followed by a live look at automated content creation.

Hands-On: Diving into 70+ Page Docs in 2 Minutes

Participants paired up or gathered three to a machine around Chromebooks pre-loaded with seven dense university IT files. 

Putting on his "teacher hat," Ricky guided the room through their first exercise. Instead of manually skimming through a massive 73-page cloud infrastructure document packed with cluster details, attendees opened the file in Google Docs and launched the Gemini side panel.

With simple, plain-language questions like "What is this file about?", the model parsed the entire 70+ page document in under two minutes. Within moments, IT leaders had extracted key infrastructure patterns, cloud resource setups, and usage trends without reading a single full page.

Connecting the Dots: Grounded AI and the "Nasi Lemak" Test

Next, the workshop moved into multi-document analysis using Gemini Notebook. Participants uploaded all seven Drive files from their folder into a single Notebook to start finding hidden connections across their dataset.

To prove how secure and accurate the platform is for Higher Education, Ricky ran a quick, humorous test: with three files selected, he typed in the prompt: *"How to cook nasi lemak?"*

Instead of pulling a recipe from the open web like standard chatbots do, Gemini Notebook politely refused, stating that no recipe existed in the uploaded university documents[4]. This demonstrated grounded AI in action—showing IT leaders that the model strictly relies on verified institutional sources, eliminating outside hallucinations and data pollution.

Uncovering $192,000 in Waste with Content Studio

With strict grounding confirmed, attendees selected specific sources—combining software license inventories with student survey results—and moved over to the Studio panel to generate executive visuals.

Within minutes, Gemini Notebook synthesized the two documents and created an infographic revealing a striking finding: the university was wasting $192,000 annually on redundant video conferencing tools (like Zoom and Webex subscriptions) that students barely used. The system also generated structured data tables summarising underutilised cloud resources, providing clear visual evidence ready to drop directly into a C-suite presentation.

Audio and written summaries powered by Gemini Notebook, a free generative AI tool grounded in your provided sources. Try Gemini Notebook for free.  You can listen to the audio overview of the session here.