Company materials for AI training: what to prepare before a workshop

Syntalith can discuss a workshop around selected team tasks before the company sends documents. Start by describing what an employee receives, what they produce, and who uses it. We can then agree on the materials needed, whether fictional examples will work, and who can explain the content. The buyer can assess a proposed task before agreeing to more extensive preparation.

Choose a difficulty worth practicing

Useful material supports work resembling the actual task. Comparing proposals requires differences that matter to the buyer. Correspondence needs the recipient's question and enough information to answer it. Exemplary finished documents show the result, but they do not always reveal the author's choices.

A program can cover explaining the goal to AI, reading sources, preparing a draft, and revising it after feedback. The instructor needs material on which those activities make sense. If the team wants practice recognizing ambiguity, the example should leave room for a question instead of supplying every answer.

CIPD's guidance on identifying learning needs includes documents, conversations, and observation, alongside confidentiality of the information collected. We choose workshop material around the agreed task and company rules.

Fictional material can preserve the real difficulty

Suppose the team wants practice comparing two service proposals. Real customer names are unnecessary for that task. A fictional example can preserve different service coverage, an ambiguous passage, and the buyer's question. If simplification makes the proposals identical, it removes the reason to analyze them. Someone familiar with purchasing can assess whether the example still calls for similar reasoning.

Removing names from a real document does not establish that the company permits its use. An authorized person on the client side assesses the content and environment. Before practice in ChatGPT Work or Claude Cowork, we agree which files may be used in the selected tool.

Material should help explain a mistake

The intended reader's expectations are useful too. The instructor can then show that a draft answers a different question, omits a meaningful difference, or needs better context. Feedback concerns work the team recognizes, giving the author a basis for an independent revision.

Someone familiar with the process explains what the information means. The learning-content author and instructor select the level of difficulty. A beginner may need a shorter collection, while a regular AI user can work with a more demanding comparison. Retaining a separate example for later independent practice requires the participant to apply the approach to new content.

If documents conflict and nobody can explain their meaning, preparation will expose that problem. The company can clarify the sources first or agree to use neutral material. Building a knowledge library or system connection requires a separate scope.

Before commissioning, we agree on example preparation, how material will be shared, and what the team will retain. Later updates and consultations also need agreement. You do not need to build a complete exercise collection before the first conversation.

Describe one task and the types of documents your team uses. Syntalith can help discuss suitable material for the program. See our pricing page for information about working together.