Glossary
Glossary
The essential terms of enterprise AI, explained simply.
AI agent
A program that pursues a goal by chaining several steps using a language model.
RAG
Retrieval-augmented generation: the model relies on passages retrieved from your documents.
Embedding
A numerical representation of a text, which makes it possible to compare content by meaning.
Chunking
Dividing a document into coherent passages before indexing.
Reranking
A second sorting pass that ranks the retrieved passages by relevance.
Hybrid search
A combination of semantic search and keyword search.
MCP
Model Context Protocol: an open standard connecting AI assistants to tools and data.
Hallucination
A statement produced by a model with no basis in its sources.
Guardrail
A rule that governs what an agent can say or do.
Evaluation
A measure of answer quality on a reference set of questions.
Go from the question to the decision
The fastest way to understand Open Intelligence is to see it answer on your own documents. Describe your need: we’ll come back to you with a tailored solution and a costed quote.