liveBook · Manning

Data agent

As explained in The Art of AI Product Development

A data agent is an AI agent that accesses, integrates, queries, and reasons over multiple organizational data sources.

A data agent unifies organizational data sources, builds a usable mental model, and answers questions through a convenient chat interface. It combines semantic search for unstructured data with Text2SQL for structured databases, then integrates retrieved evidence into an answer. section 9.4.2

A complex request can be handled as a sequence of analytical steps:

  • Retrieve product records and Q1 and Q3 2024 revenues. section 9.4.2
  • Compute changes between periods and rank the products. section 9.4.2
  • Consult product documentation and combine sources to hypothesize drivers. section 9.4.2

Revenue-growth investigation

A user asks: “What are the three products for which we had the highest revenue increase between Q1 and Q3 of 2024, and what are the possible drivers?” The agent decomposes the question, analyzes structured revenue data, and consults unstructured documentation. section 9.4.2

How The Art of AI Product Development covers it

How deep it goes
Explains source integration and question decomposition through a worked scenario, but stops short of implementation detail.
Examples
Uses a cross-source investigation of products with the highest Q1-to-Q3 2024 revenue increases.
What you can do after
Apply it via chat to investigate cross-source business questions and develop evidence-based hypotheses.
How it connects
Connects semantic search, Text2SQL, structured databases, unstructured data, and predictive foresight.

Related concepts

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