A few terms from the glossary
- AI agent
- An LLM-based program that reasons toward a goal and carries out the steps itself: it searches for data, calls tools, and produces an answer or an action.
- RAG (Retrieval-Augmented Generation)
- An approach where the model answers based on documents found in your knowledge base – with source citations. Reduces made-up answers.
- LLM
- A large language model: it understands and generates natural language. The foundation of most modern AI solutions.
- Fine-tuning
- Further training a ready-made model on your data for a specific task or answer style.
- Hallucination
- A confidently worded but incorrect answer. Reduced through RAG, scope limits, and source verification.
Who it's for
For executives, IT, and procurement who need to quickly get up to speed on the terminology before discussing an AI project – and not let anyone "sell you words".
How to use it
Keep it handy in meetings with vendors. Useful alongside the guide "AI vs RPA vs chatbot" and "12 questions for a vendor".
