# Module 2 Narration

## Opening

Open with the professional setting: a product team evaluating an NLP workflow before using it in customer-facing communication. Ask students what decision is being made, who is affected, and what evidence would be persuasive to a skeptical reviewer.

## Middle

Move through the module in four passes:

1. Define **Embeddings and semantic similarity** in the context of Natural Language Processing.
2. Walk through the lab as a proxy-data exercise, emphasizing what it can and cannot show.
3. Compare a baseline with an AI-enabled or more sophisticated alternative.
4. Translate the result into stakeholder language: recommendation, risk, mitigation, and next evidence.

## Closing

Close by returning to the module artifact: **NLP evaluation packet with task framing, retrieval/evaluation design, and deployment guardrails focused on embeddings and semantic similarity: Build a simple semantic search over course documents.**. Students should leave knowing exactly what artifact they are producing and how it will be judged.