InternshipKey terms and common mistakes
The words you will hear in AI Agents & RAG work and the mistakes that most often cost interns approvals.
Key terms
| Term | Meaning |
|---|---|
| LLM | Large Language Model |
| System prompt | Instructions setting behaviour |
| Embedding | Vector representing meaning |
| RAG | Retrieval-Augmented Generation |
| Reranker | Model re-scoring retrieved results |
| Tool calling | Model requests a function call |
| Agent | LLM that loops with tools toward a goal |
| Eval set | Test questions with expected outcomes |
Common mistakes to avoid
- Giving an agent tools with more permissions than it needs.
- Not checking whether retrieval found the right chunks before blaming the model.
- Letting the bot answer without sources when the documents do not contain the answer.
- Not limiting agent steps, so loops waste time and money.
Want to go deeper? The free AI Agents & RAG learning path has short modules, games and a certificate for each module.