Webinar RecordingData Agents in BigQuery: From Data to Answers in Natural Language
In this hands-on webinar, we will create a custom data agent in BigQuery that understands your company’s data structure, business terminology, and calculation rules, and transforms natural-language questions into ready-to-use analysis..
Who work with BigQuery and want to answer business questions faster without writing every SQL query from scratch.
Who are tired of handling daily ad hoc requests and would gladly delegate them to an agent.
Who are looking for a way to uncover insights in their data faster with the help of AI.
MARKETERS
Who want to generate analytical reports on marketing metrics independently and without SQL knowledge.
Who want to understand how to teach AI the correct business logic for accurately consolidating costs from different advertising platforms.
Who are looking for tools that provide instant answers to business questions and enable hypothesis testing without waiting for reports from analysts.
BUSINESS OWNERS
Who are looking for a simple interface for working with company data in the form of a natural conversation.
Who want to create a single agent containing business terms, calculation rules, and validated queries for the entire company.
Who want to accelerate decision-making by giving teams direct and secure access to analytics.
programs
Webinar Syllabus
PART 1. THEORY
What we will talk about
How an agent transforms natural-language questions into data analysis and SQL queries.
How to provide an agent with business context, terminology, calculation rules, and validated queries.
The capabilities, limitations, access requirements, security considerations, and cost of using data agents.
The business scenarios in which a data agent can deliver real value.
What You Will Achieve
You will understand how data agents interact with databases and learn how they differ from a standard chat interface connected to a table.
PART 2. PRACTICE
What we will talk about
An introduction to the BigQuery Studio interface and a review of the settings of a ready-made demo agent.
Access management: the roles and permissions required to use the tool.
The simplest way to interact with data: launching a basic chat directly from a data table.
What You Will Achieve
You will become familiar with the Conversational Analytics interface through ready-made solutions and learn which technical permissions are required to get started. You will also understand how the simplest format of quick chat-based queries to tables works, allowing you to retrieve initial data instantly.
What we will talk about
The key difference between a basic chat and a custom AI agent.
A step-by-step process for creating your own data agent from scratch.
How to create detailed instructions, business terminology glossaries, and validated parameterized queries.
How to use BigQuery ML’s built-in artificial intelligence capabilities for forecasting and anomaly detection.
The limitations of the technology and why tables containing raw data must never be connected.
What You Will Achieve
You will learn how to configure an AI agent from scratch based on your project’s business logic. You will also understand how built-in forecasting and anomaly detection algorithms work and how to prepare the data architecture correctly to avoid errors in the AI system’s outputs.
@makshapchuk
/maks.hapchuk