How deeply does Building Reliable AI Systems cover action execution?

Defines action execution and contrasts it with retrieval using scenarios, but stops short of code-level integration details.

Action execution

As explained in Building Reliable AI Systems

The capability of carrying out tasks by interacting with external systems rather than merely describing possible solutions.

Action execution lets an AI agent interact with external systems to carry out tasks, rather than merely describe what a person could do. It bridges the gap between an LLM’s understanding of a request and completing it through a tool such as an API or calendar. section 6.1.2 section 6.1.4

Retrieval and action serve different purposes: retrieval brings information into a response, while action changes state in an external system. An agent might retrieve flight options, then book a selected flight; retrieval-augmented generation alone cannot complete the booking. section 6.2.3 section 6.1.4

Scheduling a meeting

A standalone LLM can suggest emailing a team but cannot send an invitation. An agent can connect to a calendar, check availability, schedule the meeting, and send invites. section 6.1.2 section 6.1.4

External tools can also ground answers in current information: an agent can query a weather API instead of guessing at today’s conditions. For multi-step tasks, workflow orchestration coordinates the actions; production systems also need controls and monitoring to guard against loops that repeat tool calls without progress. section 6.2.3 section 6.0 section 6.1.4

How Building Reliable AI Systems covers it

liveBook manning.com

Read Building Reliable AI Systems on liveBook Get the book at manning.com

Related concepts

retrieval-augmented generationcontrasts with · Retrieval-augmented generation is a technique that retrieves relevant information from external sources and supplies it as context to a generative model when producing an output.
AI agentimplemented by · An AI agent is a computational system that uses artificial intelligence to interpret information, select actions, and pursue goals through interaction with an environment.
agentic RAGused by · Agentic RAG is a retrieval-augmented generation approach in which an agent dynamically decides how and when to retrieve information, often iterating or using tools before generating a response.
tool integrationproduced by · Tool integration is the process of connecting external tools to an AI system so that its model or application logic can invoke them.

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