liveBook · Manning

Vector index

As explained in Architecting Autonomous Agentic Systems

An indexed representation that supports similarity-based retrieval of documents or passages from vector embeddings.

A vector index supports similarity-based retrieval of documents or passages from embeddings. In an enterprise agent, it helps determine which evidence enters the agent’s view. That makes it a behavioral asset as well as a data asset: changes to the index or retrieval configuration can affect the agent’s interpretation. section 5.1.1 section 5.1.2

A contract amendment

A legal agent reviews a supplier’s obligations after order volumes increase and deliveries are missed. If an amendment changes those obligations, updating the source document alone is not enough. Its index status, supersession metadata and retrieval priority must also be updated, along with relationships and any context that still reflects the earlier clause. Otherwise, the agent may retrieve the original clause rather than the amendment. section 5.1.1 section 5.1.2

Manage the index alongside the knowledge base, knowledge graph, policy catalog and memory store. Assign ownership and approval for changes, and maintain versioning, lineage, access controls, reuse limits, matter isolation and retirement controls. Classification and handling constraints must continue to apply when a source becomes a retrieved passage; retrieval does not by itself establish authority to disclose or reuse it. section 5.1.1 section 5.1.2

How Architecting Autonomous Agentic Systems covers it

How deep it goes
Explains retrieval consequences and maintenance controls through a legal scenario, without detailing index implementation.
Examples
Uses a supplier agreement amendment to illustrate retrieval of an original clause versus its replacement.
What you can do after
Evaluate whether index maintenance and governance support retrieval of current contractual evidence.
How it connects
Connects vector indexes with embeddings, knowledge bases, knowledge graphs, policy catalogs and memory stores.

Also explained in

Compare how 11 books explain vector index →

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Architecting Autonomous Agentic Systems. The section links point to the parts of the book it draws on.