Concepts

Understand the core building blocks of SynapseKit.

📄️ How RAG Works

Retrieval-Augmented Generation (RAG) solves a fundamental limitation of LLMs: they can only answer questions about information in their training data.

📄️ How Agents Work

An agent is an LLM that can take actions. Instead of generating a single response, it decides what to do next, executes that action, observes the result, and repeats until the task is complete.

📄️ How Graph Workflows Work

SynapseKit's StateGraph lets you build reliable, stateful pipelines as directed acyclic graphs (DAGs). Nodes are Python functions. Edges define execution order. State flows through the graph as a shared dictionary that every node can read and write.

📄️ Embeddings and Vector Search

Embeddings are the foundation of semantic search, RAG, and recommendation systems. This guide explains what they are, how vector search works under the hood, and how to choose the right embedding model.