NotebookLM and Gemini Automate Test Creation via Google Apps Script and Forms

NotebookLM, Gemini, and Google Apps Script can be assembled into a practical pipeline for creating educational tests. Instead of manually typing out dozens of questions, a teacher or self-learner gets a semi-automated process: AI generates content, and a script immediately turns it into a ready-made Google Form.

Why This Matters

The idea here is not just about saving time. Tests are needed not only to check knowledge but also to reinforce material: when a student has to recall a topic from memory, the information sticks better. That's why short quizzes are useful both in the classroom and in self-directed learning. The problem is that with a large volume of topics, manual preparation quickly becomes tedious. You need to come up with questions, find plausible answer options, check the wording, and then transfer everything to a convenient tool for taking the test.

The stronger the knowledge of the subject, the more room for maneuver

the creative mind gets.

How the Pipeline Works

The basic scheme is built around NotebookLM, where verified sources on the topic are collected in advance. This reduces the risk of hallucinations and makes test generation more tied to the specific course materials. Next, Gemini comes into play: it receives the task of formulating questions and answer options, and then writes code for Google Apps Script based on the same dataset. The script creates a new Google Form, switches it to test mode, adds questions, marks correct answers, and assigns points.

• In NotebookLM, lecture notes, articles, and other educational materials are uploaded.
• Based on a prompt, the service generates a set of questions on a chosen topic.
• Gemini converts this set into code for Google Apps Script.
• The script automatically creates a Google Form and fills it with questions.
• The finished form can be immediately adjusted and sent to students.

In the initial phase, the process was still partially manual. You had to separately contact NotebookLM, then transfer the results to Gemini, get the script, insert it into the editor, and after running it, move the created form to the right folder in Google Drive. But even this version already saved hours compared to fully manual test assembly, especially if it involved series of similar quizzes on basic topics.

What Improved

The next improvement was a tighter integration of Google services. The author discovered that a NotebookLM notebook can be attached directly to a Gemini chat, which means removing one extra step. After that, a single request was enough: ask Gemini to use the materials from the notebook, generate questions, and immediately provide ready code for Google Apps Script to create the form.

In a practical scenario, this noticeably speeds up the work: copy-pasting becomes less necessary, and the path from lesson topic to finished test shrinks to just a few minutes. However, complete automation wasn't achieved. The idea of a web app that was supposed to turn the process into almost one button hit a wall with templates and parsing the question structure.

This option turned out to be interesting but not particularly useful in real work. Additionally, the author emphasizes an important boundary: for simple tests on basic knowledge, automation works great, but complex and finely tuned tasks still require manual editing and final human review.

What This Means

This case shows that useful automation in education no longer requires separate development or a paid stack. It's enough to assemble a clear low-code process from NotebookLM, Gemini, Google Forms, and Apps Script. For teachers, it's a way to produce review materials faster, and for self-learning, it's a convenient mechanism to regularly drill yourself on a topic without spending time on routine instead of actual learning.