AI & business tech glossary
Plain-English definitions of the terms we use across Corewire. No jargon for its own sake.
A quick reference to the concepts behind our reporting. Each entry is short on purpose; follow the links to see the ideas in action.
Agentic AI
AI that does not just answer, it acts. An agentic system breaks a goal into steps, uses tools (software, search, code), checks its own progress, and keeps going with limited supervision. A chatbot replies to one message; an agent works a task from start to finish. Related: our Artificial Intelligence coverage.
Large Language Model (LLM)
The engine behind tools like ChatGPT, Claude, and Gemini. An LLM is trained on huge amounts of text to predict the next piece of a sentence, which lets it write, summarise, translate, and answer questions. It does not store facts like a database; it generates likely text, which is why it can be fluent and wrong at the same time. Related: Tokens and Retrieval-Augmented Generation.
Retrieval-Augmented Generation (RAG)
A way to make an LLM answer from specific, trusted documents instead of only its training. The system first retrieves relevant text (your files, a database, the web), then asks the model to answer using it. That reduces made-up answers and lets a model cite sources it never trained on. Related: Large Language Model.
Tokens
The units an LLM reads and writes. A token is roughly three-quarters of a word, so a short sentence is a dozen or so tokens. Models have a token limit (the “context window”) and usage is usually billed per token, so token count drives both how much a model can consider at once and what it costs. Related: Large Language Model.
See also
- Prompt templates: ready-to-use prompt structures.
- Cheat sheets: quick references for working with AI.
- How to use Prompt Forge: turn a rough idea into a clean prompt.