Introducing Conversational Analytics in BigQuery

Businesses want to move quickly and make informed decisions, but the explosion of data in today’s organizations often can leave knowledge teams buried and business users waiting in lengthy queues for the data insights they need. AI agents promise to fundamentally change this relationship, empowering users to move faster from data to action.

Today, we are unveiling Conversational Analytics in BigQuery in preview. This new offering allows users to analyze data using natural language, breaking down the knowledge walls and time sinks that have long been the norm. Following the general availability of Conversational Analytics in Looker, this integration brings a sophisticated AI-powered reasoning engine directly into BigQuery Studio.

Data insights for data professionals, made conversational

Conversational Analytics in BigQuery is more than a simple chatbot. It’s an intelligent agent that leverages the latest Gemini models to generate, execute, and visualize answers grounded in your specific business context directly in BigQuery’s secure and scalable environment. 

With Conversational Analytics in BigQuery, technical and business teams can build and deploy intelligent agents at the source, leveraging their existing data and metadata for rapid innovation. Now you can create context- and business-aware agents right where the data lives, so that all users get smart insights, coached by trusted analysts, without ever having to wait for answers in a queue or learn SQL!

From question to trusted answer in seconds

Unlike simple data tools, Conversational Analytics in BigQuery uses your business metadata and production logic to build trust between the user and their data. 

When a user asks a question, the agent employs a multi-stage workflow to guarantee that every response is both precise and contextually relevant. These AI-driven insights provide users with comprehensive analysis through summarized answers, raw data results, and visualizations, supplemented by follow-up questions to facilitate further investigation.

Conversational Analytics in BigQuery is characterized by:

• Being grounded in reality: By leveraging your BigQuery schema, metadata, and custom instructions, the agent ensures SQL generation is based on internal logic rather than generic assumptions.

• Verified queries and trusted logic: To maintain consistency with production metrics, you can ground the agent in verified queries and User Defined Functions (UDFs). This leverages your team’s enterprise-ready assets so you don’t have to reinvent the wheel.

• Transparent logic and summarization: To give you confidence, the agent surfaces its “thinking process” and the generated SQL behind every answer. It then synthesizes the insights it gained across thousands of rows and provides a concise executive summary explaining the “why” behind the numbers.

• Security and governance by design: Users only access data they are authorized to view, with every query logged for auditing within the BigQuery compliance framework.

Beyond queries: Predict what’s next in seconds

While most tools report on data retrospectively, Conversational Analytics in BigQuery transforms the experience from retrospective to predictive. By leveraging BigQuery AI, agents can forecast outcomes and uncover hidden patterns using simple language.

Behind the scenes, the agent uses functions like AI.FORECAST to predict trends or AI.DETECT_ANOMALIES to surface outliers in real-time. This allows any user to perform advanced predictive analytics in seconds, without leaving the chat interface. The agent leverages generative AI to distill millions of rows into a clear story, quickly making insights that are contextual and easy to share.

Unlocking the value of unstructured data

With Conversational Analytics in BigQuery, you’re not limited to data in rows and columns. The agent can reason across unstructured data, such as images stored in BigQuery object tables. This lets you query your entire data estate from a single interface, transforming previously inaccessible information into actionable insights with no manual processing.

Bring agents to life

We built Conversational Analytics in BigQuery to let you transform raw data into active, intelligent agents with minimal effort. By simply connecting your tables and adding specific business instructions and metadata, you can move beyond manual queries to automated insights. BigQuery’s assisted authoring helps you create quality agents quickly, which can then be shared across Looker Studio Pro the BigQuery UI.

The agents can also be integrated into their own custom apps and existing agentic ecosystems via the API and ADK tools.

Transform your data analytics today

If you’re ready to tackle the data analytics bottleneck, you can access the preview of Conversational Analytics in BigQuery starting today. For more information, including a deep dive into best practices for context-based grounding and API integration, please refer to our documentation or learn more about Google Cloud’s AI agents for data analytics.