The Revision Engine: How to scale student support without burnout.

Every teacher knows the intense pressure of the pre-exam season. Students are anxious, past papers are finite, and the familiar request for "extra practice questions and video links" starts flooding your inbox.

Traditionally, building a comprehensive, curated revision guide is a weekend killer. It means hunting down specific YouTube tutorials, verifying that they use the correct graphic display calculator (GDC) models, finding aligned practice questions, and organizing it all into a coherent document.

This year, I decided to shift my workflow. Instead of burning out on administrative curation, I used Gemini to build a Revision Engine for my Applications and Interpretation SL (AISL) students.

By using AI to handle the heavy lifting of sourcing and structuring, I created a targeted revision tracker that ultimately supported 50 students across our cohort, generating hundreds of hours of productive, independent study time.

But the most surprising result? Automating the logistics didn't distance me from my students. It drove them into my classroom.

Here is the exact recipe, prompt, and quality-control checklist I used, so you can build this sustainable practice for your own classroom.

The Secret: Context Curation

If you simply ask an AI for "IB Math revision links," you will get generic, frequently broken, or completely misaligned results. To make an AI work effectively as your research assistant, you must feed it your specific instructional parameters.

Before running the prompt below, upload your AISL Syllabus Outline, your Internal Teaching Overview/Sequence, and your Calculator Guide directly into the AI interface.

The Prompt Recipe

Feel free to copy and paste this prompt template to generate your initial structure:

Role: You are an expert IB Diploma Programme Mathematics Specialist and data organizer. Task: Create a structured raw CSV/Markdown table to help my DP Applications and Interpretation SL (AISL) Year 1 students track their revision for their upcoming exams. Table Schema Requirements:Organize the table with the exact columns:[Topic], [Unit Ref], [Suggested YouTube Video Title Search], [Google Forms Quiz], [Practice Questions], [Solutions], [Complete?], [RAG Rating] Data Instructions:Topic & Unit Ref: Map out the foundational Year 1 AISL topics (Number & Algebra, Functions, Geometry & Trig, Statistics & Probability) and cross-reference them with their official IB syllabus sub-topic unit references (e.g., Scientific Form to 1.1, Voronoi Diagrams to 3.6, Chi-Square to 4.11). Suggested YouTube Video: Generate optimized YouTube video search queries for each topic that are highly likely to exist, specifically appending the tag [IB Math AI SL/HL] to ensure syllabus alignment. Placeholders: For columns like Google Forms Quiz, Practice Questions, and Solutions, populate them with text placeholders like [Insert Google Form URL] or [Insert Textbook Page/Link] so I can easily find and replace them with my localized school resources. Default Status: Set Complete? to FALSE and RAG Rating to a default empty space or a placeholder letter (r, a, g) for students to color-code later. Output Format: Provide the response strictly as a clean, copy-pasteable markdown table.

The "Human in the Loop" Cleaning Checklist

Let’s be realistic: AI is a phenomenal structural framework builder, but it is not a certified IB examiner. It will generate dead URLs if you ask it for links, or it might suggest a video that doesn't quite hit the mark.

However, correcting an existing, 80% accurate skeleton table is a massive time-saver compared to staring at a blank spreadsheet. Once Gemini outputted the raw table, I spent a few hours running it through this 3-Point Curation Checklist before distributing it to the cohort:

1. The Textbook Anchor

Instead of relying on random web links, I anchored the spreadsheet directly to our physical school curriculum assets. Where the AI left placeholders for practice questions, I went in and added explicit page and exercise numbers from our textbook (e.g., Exercise 3B, Page 94 in the Hodder textbook). Giving students exact parameters means zero friction when they sit down to study at night.

2. The Internal Asset Swap (The "Form & Cloud" Pass)

I replaced the placeholder cells in the Quiz and Solutions columns with our actual classroom data infrastructure. I dropped in live links to our diagnostic Google Forms Quizzes and pinned direct paths to our cloud-hosted solution keys (like a Dropbox PDF for polynomial modeling solutions). This turned a static document into a live link between independent study and teacher analytics.

3. The Technology Filter (The GDC Pass)

In AISL, knowing how to leverage technology is the syllabus skill. I audited the video suggestions for topics like Systems of Linear Equations (1.8) or Key Features & Intersections (2.4) to ensure they specifically highlighted GDC Tips. If a video didn't show the exact calculator model our students use (like the TI-84 Plus CE), I swapped it for one that did.

The "High-Tech, High-Touch" Paradox

When we talk about "sustainable digital practices" on this blog, we aren't trying to find a shortcut to leave school earlier on a Friday. We are trying to protect our cognitive and emotional capacity so that we can be fully present for our students when it matters most.

By letting the AI handle the mechanical work of organizing the revision engine, and pairing that with automated "schedule send" outreach emails to check in on students during study leave, something beautiful happened.

The technology didn't distance me from the cohort. It acted as an open invitation.

Because the administrative logistics were entirely taken care of, my classroom space became a hub for real-life, in-person connection. The spreadsheet gave students a common language; they could look at a topic they had marked as "Red" (r), click the video, check the textbook, and if they still didn't get it, they knew exactly where to go.

Those scheduled emails and structured lists broke down the barrier of asking for help. My room quickly turned into a space for quiet reassurance, pre-exam debriefs, and check-ins over a cup of tea for students who I knew would otherwise struggle in isolation.

The True ROI

Three hours after our students sat their final mathematics papers, the real data came in. From my own small class of 12, half reached out immediately—three via direct, heartfelt emails and another three hunting me down in person to say thank you and talk through their experience.

But the real validation of this sustainable ecosystem? Another eight students whom I don't even teach did the exact same thing. They had access to the Revision Engine, felt entirely supported by the structure, and sought out that human connection to celebrate the finish line.

Technology should never replace the relational heart of teaching. But when used intentionally, it strips away the noise so that what remains is purely human: connection, understanding, and community.

Some lovely feedback from a student