---
title: "Embedding"
date: 2025-11-04
author: "Anton \"JD\" Titov"
---

# Embedding

An **Embedding** is a numerical representation (a vector) of data – typically text – that mathematically encodes the content’s meaning and context. It is the language that **[AI](https://stage.spiralscout.com/glossary/artifical-intelligence)** uses to understand similarity. Phrases with similar meaning will have embeddings that are numerically close together.

Spiral Scout utilizes embeddings to power **[Semantic Search](https://stage.spiralscout.com/glossary/semantic-search)** and **[RAG](https://stage.spiralscout.com/glossary/retrieval-augmented-generation)** capabilities, enabling agents to quickly retrieve relevant knowledge based on conceptual meaning rather than just keyword matches.