NotebookLM for Training
& Course Creation

Google NotebookLM is a free AI research assistant that works exclusively with sources you provide. Upload PDFs, Google Docs, websites, YouTube videos, or audio files. Ask questions. Get answers grounded only in your content. No hallucination from general training data. No internet search mixed in.

Why L&D professionals should care

Course creation starts with research: SME interviews, policy documents, technical manuals, existing training materials. NotebookLM turns that pile of source material into a queryable knowledge base. Instead of reading 200 pages to find the three key concepts for Module 3, you ask NotebookLM and it tells you - with citations back to the exact source.

Every NotebookLM project starts with sources. You can upload up to 50 sources per notebook, and each source becomes a "grounding" reference. NotebookLM answers only from these. This is the foundation of everything you'll build.

Supported source types

Organising sources for a course project

Think of each notebook as one course or one major module. Group your sources intentionally:

COURSE SOURCE STRUCTURE (one notebook per course)

Primary sources (add first - these anchor all responses):
- Course brief / learning objectives document
- SME interview transcripts (audio or text)
- Existing training materials being updated

Supporting sources:
- Policy documents / procedures / standards
- Research papers / industry guidelines
- Competitor course outlines (for gap analysis)

Reference sources:
- Relevant YouTube tutorials (NotebookLM transcribes these)
- Industry website URLs
- Organisational style guides / brand docs

This is the feature that makes people's jaws drop. NotebookLM can generate a 10-15 minute podcast-style audio overview of your sources. Two AI hosts discuss the material conversationally, covering key points, explaining concepts, and even debating nuances. It sounds remarkably natural.

How to guide the conversation

You are not stuck with a generic overview. Before generating, you can add custom instructions that shape what the hosts focus on. This is critical for training use cases.

The Notebook Guide is a set of persistent instructions you write that shape every response NotebookLM gives within that notebook. Think of it as the system prompt for your course development project. Every time you ask a question, NotebookLM follows these instructions automatically.

The L&D Notebook Guide template

Copy this into the Notebook Guide section of your course development notebook. It transforms every response into instructional design output.

You are an experienced instructional designer helping me develop a training course.

For every response:
1. Structure content as: Learning Objective > Key Concepts > Supporting Details > Application Activity
2. Write learning objectives using action verbs (analyse, demonstrate, apply - not 'understand' or 'know')
3. Identify where scenario-based assessment would be most effective
4. Flag any content gaps where the sources don't provide enough information
5. Suggest which content is essential vs nice-to-have for time-constrained learners
6. Always cite which source document each piece of information comes from

Audience: [describe your learner audience here]
Course format: [eLearning / blended / instructor-led / self-paced]
Approximate duration target: [X hours/modules]

This is the practical workflow. Five steps that take you from a pile of source documents to a structured course outline with learning objectives and assessment ideas. Each step builds on the last.

Analyse all sources in this notebook. What are the key topics and concepts that should be covered in a training course on this subject? Group them into logical themes and indicate which sources support each topic.

Based on the key topics identified, create a module outline for a [duration] [format] course. For each module include: title, duration estimate, 2-3 key learning points, and prerequisite knowledge from earlier modules. Order modules from foundational to advanced.

Write 3-5 learning objectives for Module [X]: [module title]. Each objective should: use an action verb, be measurable, and be directly supported by specific content in the source documents. After each objective, note which source and section supports it.

Assessment generation is where most AI tools fail. They produce generic quiz questions that test recall of definitions rather than application of skills. NotebookLM is different because it generates assessments grounded in your actual training content. The questions, scenarios, and case studies reference real situations from your source material.

Three assessment types to generate

Generate 5 multiple-choice questions that test APPLICATION of the concepts in [module/topic]. Each question should present a realistic workplace scenario where the learner must decide the correct action. Include 4 options per question with plausible distractors. After each correct answer, explain WHY it is correct with a citation to the relevant source.

Create a scenario-based case study that tests the learner's ability to apply [skill/concept]. The scenario should: (1) describe a realistic situation based on the information in the sources, (2) present 3 decision points where the learner must choose an action, (3) include consequences for each choice, and (4) end with a debrief explaining the optimal path with source citations.

Based on the information in [Source name/number], write a case study where a learner must apply the key procedures described. Include: background context, the specific challenge, 3 questions the learner must answer, and a model answer for each with direct quotes from the source material.

Five ready-to-use prompts that cover the full course development workflow. Copy them. Paste them into your NotebookLM notebook. Adjust the bracketed sections for your project. Use them this week.

Analyse all sources in this notebook. For each source, provide: (1) a one-sentence summary, (2) the key topics covered, (3) the most important facts or procedures, and (4) how this source relates to the others. Then identify any gaps - topics that seem important but are not well-covered by any source.

Create a complete course outline for a [duration] [format] course based on the sources in this notebook. Target audience: [describe learners]. For each module include: title, estimated duration, 3-5 key learning points, suggested activities, and prerequisite knowledge. Order from foundational concepts to advanced application. Flag any modules that need additional source material.

Write 3-5 learning objectives for a module on [topic]. Requirements: (1) each must use a measurable action verb, (2) each must be achievable within the module's timeframe, (3) each must be directly tied to content in the source documents - cite which source supports each objective, (4) include one objective that tests application in a realistic scenario, not just recall. Format: 'By the end of this module, the learner will be able to [action verb] [specific skill/knowledge] [context/condition].'

Create an assessment for [module/topic] with: (1) 3 scenario-based multiple choice questions where the learner must apply what they learned to a realistic situation, (2) 1 case study with 3 decision points, and (3) 2 reflection questions for self-assessment. For each question, cite the source material that supports the correct answer. Make distractors plausible but clearly incorrect based on the source content.

Focus the conversation on the practical application of [topic/module]. Explain the key concepts as if preparing someone for a hands-on workshop. Cover: (1) the 3 most important things to understand before the session, (2) common mistakes people make and how to avoid them, (3) real-world examples from the source material. Keep the tone conversational but professional. End with 3 questions the listener should be able to answer after listening.

You now know how to use Google NotebookLM to turn scattered source material into structured, grounded training content. Here is what you covered: