AI can turn documents into an e-learning course, but what happens when the most important knowledge isn’t in those documents? Experienced employees often rely on judgement, unwritten rules, and exceptions that standard procedures don’t explain. Before creating a course with AI, e-learning experts need to capture that tacit knowledge through focused interviews, document key decisions, and validate them with subject-matter experts. Here’s how to prepare reliable source materials that AI can use to draft a course.
Why E-Learning Experts Need to Capture Tacit Knowledge
Company procedures explain how tasks should be performed, but they don’t always show how experienced employees handle unusual situations. These decisions often depend on tacit knowledge: practical expertise developed through experience that may be difficult to put into words.
For e-learning experts, the challenge is identifying which parts of that knowledge learners need. Start by reviewing existing training materials, then ask subject-matter experts to explain decisions that depend on context, exceptions, or competing priorities.
Some of these decisions can become practical examples or scenario-based exercises. Others may require discussion, demonstration, or guidance from an experienced colleague. The goal is to determine what can be taught through a course and what requires additional support.
How to Capture Tacit Knowledge Through Expert Interviews
Instead of asking experienced employees to describe their entire job, focus on specific decisions they make at work. Ask what information they consider, why they choose a particular action, and what might lead them to act differently.
For example, ask an expert to describe a situation where the standard procedure wasn’t enough. What made the case unusual? How did they decide what to do? Their answers can reveal practical rules and exceptions that deserve a place in the training course.
According to APQC’s guidance on capturing tacit knowledge, structured interviews, knowledge mapping, and practical examples can help organisations document expertise that might otherwise be difficult to share. However, not all tacit knowledge can be fully captured in writing.
During the interview, record not only what the expert recommends but also why. If they describe an exception, establish when it applies and whether the standard procedure remains valid in other situations. This helps prevent learners from treating an exceptional case as a general rule.
For more complex decisions, create a simple decision map showing the available options and the conditions behind each choice. Ask the expert to review it, correct any inaccuracies, and identify unresolved questions before the information becomes part of the course.
How to Prepare Expert Knowledge for AI Course Creation
After interviewing the subject-matter expert, organise the findings into materials that can be reviewed before AI course creation begins. A simple decision map should explain when a particular action is appropriate, why it is recommended, and what exceptions apply.
Include practical examples that illustrate these decisions, along with any questions that remain unresolved. Ask the expert to check the accuracy of the information and confirm which parts are ready to be used in training. Keep unverified assumptions out of the approved materials.
|
Expert prompt |
Reviewable asset |
Expert approval check |
|---|---|---|
|
When do you choose a different action? |
A decision rule with its conditions and a worked exception. |
Confirm the conditions and flag unresolved cases. |
Ask the expert to review the content for accuracy, not just clarity. Make sure each decision rule includes the relevant conditions and exceptions, and resolve any disagreements before approving the materials for AI processing.
Once the materials are approved, the team can move on to course development. Our guide to creating an online course with AI from existing materials explains how to turn those materials into structured e-learning content.
Use AI to Draft From Expert-Reviewed Materials
AI4E-learning is an AI-powered e-learning authoring tool developed by TTMS. It uses supplied materials, including DOCX, PDF, PPTX, MP3, and MP4 files, to create editable course drafts. The application supports course structuring, content editing, approval, and SCORM export.
Once the expert knowledge has been collected and approved, the L&D team can use AI4E-learning to prepare an initial course draft. The expert should then check whether the content accurately reflects the documented decisions, including the conditions and exceptions that influence them.
Pay particular attention to scenarios and exercises. If AI introduces an assumption that isn’t supported by the source materials, consult the expert rather than treating the generated explanation as fact. The L&D team should also verify that the exercises teach the intended skills and reflect situations learners may encounter at work.
How to Ensure Content Accuracy and Data Privacy in AI-Generated Courses
Before publishing an AI-generated course, the subject-matter expert should review its content for accuracy. This includes checking whether decision rules reflect actual procedures, exceptions are explained correctly, and scenarios don’t introduce misleading assumptions. The L&D team should separately assess whether the exercises help learners practise the intended skills in realistic situations.
Human oversight is also emphasised in UNESCO’s guidance on generative AI in education and research, which highlights the importance of content validation and data privacy. Although these recommendations address education broadly, they provide useful principles for organisations developing AI-assisted training.
Data protection becomes particularly important when source materials include internal procedures, employee interviews, or confidential business information. Before sharing these materials with an AI vendor, determine which documents can be used, who is authorised to access them, and whether any sensitive details should be removed or anonymised.
The L&D team should also ask the vendor how uploaded information is stored, processed, and protected, including whether it is retained after course creation. These arrangements should be confirmed against the organisation’s security and privacy requirements before any sensitive materials are transferred.
What to Ask an E-Learning Vendor Before Creating a Course With AI
When choosing an e-learning vendor, look beyond their ability to generate courses from existing documents. Find out how they approach situations where important expert knowledge hasn’t yet been documented.
Ask whether the proposed process includes interviews with subject-matter experts, identifying exceptions to standard procedures, and reviewing the collected information before AI course creation begins. If these activities aren’t included, establish who in your organisation will be responsible for them.
It’s equally important to understand what happens after AI generates the first draft. Who checks whether the course reflects the expert’s knowledge? How are unclear or conflicting instructions resolved? A demonstration of automated course creation won’t necessarily answer these questions.
At TTMS, our e-learning services include instructional design, course planning, and working with expert-provided materials. The exact scope of knowledge gathering and expert involvement should be agreed upon before the project begins, based on your organisation’s needs. TTMS Malaysia is a Registered Training Provider with HRD Corp, so Malaysian employers can fund TTMS training from their HRD Corp levy.
If you’re planning to create training materials with AI, talk to our team about how to prepare expert knowledge for course development. Learn more about our e-learning services.
FAQ: Settle Knowledge-Capture Decisions
What if two experts describe the same exception differently?
Document both explanations and identify the circumstances in which each applies. Ask the experts to clarify any differences and agree on the correct approach before including it in the training materials.
Can every expert judgement be turned into a document?
Not always. Some tacit knowledge is difficult to document because it depends on experience and context. Practical examples, expert discussions, and peer learning can help transfer knowledge that written instructions alone cannot fully explain.
Who should approve an AI-generated scenario before learners see it?
A subject-matter expert should verify the accuracy of the scenario, including relevant decisions and exceptions. The L&D team should then check whether it supports the intended learning objectives. Both reviews should be completed before publication.
What should the team check before sharing expert materials with an AI vendor?
Confirm which documents can be shared, who will have access to them, and how the vendor processes, stores, and protects the data. Make sure these arrangements meet your organisation’s privacy and security requirements before uploading any sensitive materials.
