AI data analytics agents
Supercharge data analytics workflows with agentic AI
Simplify workflows, empower your teams, and accelerate insights with intelligent data analytics agents.
Overview
How can AI agents be used for data analytics?
AI agents help data teams automate repetitive tasks like data cleaning and labeling, and business users analyze data and predict outcomes using natural language. This frees up various teams from mundane work, allowing them to focus on higher-value strategic initiatives. The result is faster insights, quicker innovation, and more efficient scaling of AI across the organization.
Who can use AI agents for data analytics?
AI agents act as powerful allies across the entire data organization:
- Data engineers: Automate pipeline creation and maintenance using natural language prompts
- Data scientists: Streamline data wrangling, model evaluation, and feature engineering
- Analysts and business users: Gain instant insights and generate visualizations by asking questions in plain English, removing the need for specialized coding
Why build AI data analytics agents on Google Cloud?
- Unified AI-native foundation: It explains how Google Cloud's platform isn't just a collection of siloed tools but a single, integrated environment that removes the divide between analytical and operational data
- Real-time business context: This integration allows agents to have a comprehensive, real-time understanding of the business, which is critical for their effectiveness
- Contextual grounding: It emphasizes that agents are grounded in your specific data environment—using metadata, schemas, and lineage to ensure accuracy and alignment with business definitions
How It Works
Automated data engineering
Automate complex and time-consuming data engineering tasks
The Data Engineering Agent in BigQuery is an intelligent assistant powered by Gemini that moves beyond simple code completion to provide end-to-end task automation. It's grounded in your specific data environment—it uses metadata from Knowledge Catalog (formerly Dataplex) to understand your schemas, lineage, and business definitions. It can autonomously handle the entire data life cycle: from discovering relevant datasets and generating complex SQL or PySpark transformations to orchestrating those jobs through Dataform or Cloud Composer. By automating the "toil" of data engineering—such as fixing broken pipelines, documenting legacy code, or migrating queries from older data warehouses—it transforms the role of the data engineer from a manual coder into an architect who supervises AI-driven workflows.
“The agent provides solutions that enable us to explore new development approaches, showing strong potential to address complex data engineering tasks. It demonstrates an impressive ability to correctly interpret our requirements, even for sophisticated data modeling tasks like creating SCD Type 2 dimensions. In its current state, it already delivers value in automating maintenance and small optimizations, and we believe it has the foundation to become a truly distinctive tool in the future.”- Fernando Calo, Lead Data Engineer at the Spanish-language news and entertainment group PRISA
“During the migration journey to a Dataform environment, the Data Engineer Agent successfully replicated all existing data and transformations scripts with 100% automation and zero manual intervention. This achievement resulted in a 90% reduction in the time typically required for manual ETL migration, significantly accelerating the transition." - Chris Benfield, Head of Engineering, Vodafone
“Process documentation is often a tedious task for developers, but with the Dataform Data Engineering Agent this effort is fully automated. The agent was able to accurately generate documentation directly from our Dataform project files, following the standards and styles we defined. This allowed us to keep our documentation consistently up to date as changes were introduced, enabling zero manual intervention in our documentation workflow. It proved to be a tool with significant potential.” - Maximiliano Morales, Data Engineer at a leading telco in Argentina
How-tos
Automate complex and time-consuming data engineering tasks
The Data Engineering Agent in BigQuery is an intelligent assistant powered by Gemini that moves beyond simple code completion to provide end-to-end task automation. It's grounded in your specific data environment—it uses metadata from Knowledge Catalog (formerly Dataplex) to understand your schemas, lineage, and business definitions. It can autonomously handle the entire data life cycle: from discovering relevant datasets and generating complex SQL or PySpark transformations to orchestrating those jobs through Dataform or Cloud Composer. By automating the "toil" of data engineering—such as fixing broken pipelines, documenting legacy code, or migrating queries from older data warehouses—it transforms the role of the data engineer from a manual coder into an architect who supervises AI-driven workflows.
Customer examples
“The agent provides solutions that enable us to explore new development approaches, showing strong potential to address complex data engineering tasks. It demonstrates an impressive ability to correctly interpret our requirements, even for sophisticated data modeling tasks like creating SCD Type 2 dimensions. In its current state, it already delivers value in automating maintenance and small optimizations, and we believe it has the foundation to become a truly distinctive tool in the future.”- Fernando Calo, Lead Data Engineer at the Spanish-language news and entertainment group PRISA
“During the migration journey to a Dataform environment, the Data Engineer Agent successfully replicated all existing data and transformations scripts with 100% automation and zero manual intervention. This achievement resulted in a 90% reduction in the time typically required for manual ETL migration, significantly accelerating the transition." - Chris Benfield, Head of Engineering, Vodafone
“Process documentation is often a tedious task for developers, but with the Dataform Data Engineering Agent this effort is fully automated. The agent was able to accurately generate documentation directly from our Dataform project files, following the standards and styles we defined. This allowed us to keep our documentation consistently up to date as changes were introduced, enabling zero manual intervention in our documentation workflow. It proved to be a tool with significant potential.” - Maximiliano Morales, Data Engineer at a leading telco in Argentina
Accelerated data science
Accelerate data exploration to model evaluation and MLOps
The Data Science Agent in BigQuery accelerates data science development with agentic capabilities that facilitate data exploration, transformation, and ML modeling.
With a simple prompt, the agent generates a detailed plan covering all aspects of data science modeling from data loading, exploration, cleaning, visualization, feature engineering, data splitting, model training/optimization and evaluation. If the agent makes an error, it can autocorrect and generate new code to rectify it. You maintain full control, with the ability to approve each step and make manual edits if desired.
The agent also has full contextual awareness of your notebook, understanding existing code, outputs, and variables to provide tailored code for each step of the plan, allowing you to make iterative changes to your existing code.
"The Data Science Agent has been a game-changer for our data science team. It streamlines our workflow by taking simple, natural language instructions and translating them into multi-step data science code, which it then executes. We longer have to start from scratch with the code. Features like code completion, error fixing, and natural language-based visualization have shown the team how AI can be an accelerator for data scientists.” - Lorraine Zheng, Data Scientist at Snap Inc.
How-tos
Accelerate data exploration to model evaluation and MLOps
The Data Science Agent in BigQuery accelerates data science development with agentic capabilities that facilitate data exploration, transformation, and ML modeling.
With a simple prompt, the agent generates a detailed plan covering all aspects of data science modeling from data loading, exploration, cleaning, visualization, feature engineering, data splitting, model training/optimization and evaluation. If the agent makes an error, it can autocorrect and generate new code to rectify it. You maintain full control, with the ability to approve each step and make manual edits if desired.
The agent also has full contextual awareness of your notebook, understanding existing code, outputs, and variables to provide tailored code for each step of the plan, allowing you to make iterative changes to your existing code.
Customer examples
"The Data Science Agent has been a game-changer for our data science team. It streamlines our workflow by taking simple, natural language instructions and translating them into multi-step data science code, which it then executes. We longer have to start from scratch with the code. Features like code completion, error fixing, and natural language-based visualization have shown the team how AI can be an accelerator for data scientists.” - Lorraine Zheng, Data Scientist at Snap Inc.
Conversational analytics in BigQuery
Make BigQuery insights accessible to data teams
Conversational Analytics in BigQuery is a sophisticated AI-powered reasoning engine that eliminates the "analytics bottleneck" by empowering data teams to bridge the gap between business questions and trusted answers through intuitive natural language interaction.
Powered by the latest Gemini models, this agent moves beyond simple translation by grounding its logic in your organization's existing data assets—specifically leveraging metadata, table and column descriptions, business glossaries, and user defined functions—to ensure every generated result aligns perfectly with your internal business definitions. Moving beyond simple reporting, the agent uses BigQuery AI to project future outcomes and interprets unstructured data like images within object tables to turn hidden information into actionable intelligence.
Tip: AI agents are most effective when they have a unified, real-time understanding of the business, removing the divide between analytical and operational data.
"With BigQuery's Conversational Analytics, we've further accelerated how our teams interact with data at Pet Circle. By allowing our teams to ask complex data questions in natural language, we’ve drastically reduced our time-to-insight. It empowers our data teams to create agents for non-technical teams, enabling them to make faster, data-driven decisions that ultimately help us deliver a better experience for pet parents." - Alistair Venn, CEO of Pet Circle
How-tos
Make BigQuery insights accessible to data teams
Conversational Analytics in BigQuery is a sophisticated AI-powered reasoning engine that eliminates the "analytics bottleneck" by empowering data teams to bridge the gap between business questions and trusted answers through intuitive natural language interaction.
Powered by the latest Gemini models, this agent moves beyond simple translation by grounding its logic in your organization's existing data assets—specifically leveraging metadata, table and column descriptions, business glossaries, and user defined functions—to ensure every generated result aligns perfectly with your internal business definitions. Moving beyond simple reporting, the agent uses BigQuery AI to project future outcomes and interprets unstructured data like images within object tables to turn hidden information into actionable intelligence.
Tip: AI agents are most effective when they have a unified, real-time understanding of the business, removing the divide between analytical and operational data.
Customer examples
"With BigQuery's Conversational Analytics, we've further accelerated how our teams interact with data at Pet Circle. By allowing our teams to ask complex data questions in natural language, we’ve drastically reduced our time-to-insight. It empowers our data teams to create agents for non-technical teams, enabling them to make faster, data-driven decisions that ultimately help us deliver a better experience for pet parents." - Alistair Venn, CEO of Pet Circle
Conversational analytics in Looker
Chat with your data
Conversational Analytics in Looker simplifies Business Intelligence by enabling business users to find answers using natural language. This reduces the burden on data analysts and facilitates faster, more confident decision-making. Business users can ask direct questions about product performance or traffic trends without needing to understand complex field names.
Beyond simple querying, it provides a comprehensive life cycle management framework, incorporating enterprise-grade security and user management directly into the consumption layer. The universal semantic layer of Looker ensures that metrics like revenue and churn remain consistent across the company by creating a central hub for data context, definitions, and relationships.
"Effective conversational analytics starts with a unified, audited data layer. If teams aren't speaking the same data language, AI systems can't reliably interpret queries or surface accurate insights." - John Pettit Chief Technology Officer, Promevo
“Our vision is for customers not only to see what happened, but to have a conversation with their data and receive intelligent recommendations inside IRIS Fleet and our other products. We believe the real opportunity is just beginning.” - Gerardo Ortiz, Head of Product and Digital Transformation, Métrica Móvil.
How-tos
Chat with your data
Conversational Analytics in Looker simplifies Business Intelligence by enabling business users to find answers using natural language. This reduces the burden on data analysts and facilitates faster, more confident decision-making. Business users can ask direct questions about product performance or traffic trends without needing to understand complex field names.
Beyond simple querying, it provides a comprehensive life cycle management framework, incorporating enterprise-grade security and user management directly into the consumption layer. The universal semantic layer of Looker ensures that metrics like revenue and churn remain consistent across the company by creating a central hub for data context, definitions, and relationships.
Customer examples
"Effective conversational analytics starts with a unified, audited data layer. If teams aren't speaking the same data language, AI systems can't reliably interpret queries or surface accurate insights." - John Pettit Chief Technology Officer, Promevo
“Our vision is for customers not only to see what happened, but to have a conversation with their data and receive intelligent recommendations inside IRIS Fleet and our other products. We believe the real opportunity is just beginning.” - Gerardo Ortiz, Head of Product and Digital Transformation, Métrica Móvil.
Conversational analytics API
Integrate agentic workflows into your applications
The Conversational Analytics API lets developers embed natural-language query functionality in custom applications, internal tools, or workflows, all backed by trusted data access and scalable, reliable data modeling. It’s the same API that powers the out-of-the-box conversational experiences in Looker and BigQuery.
The Conversational Analytics API lets you build custom data experiences that provide data, chart, and text answers while leveraging Looker's trusted semantic model for accuracy or providing critical business and data context to agents in BigQuery. You can embed this functionality to create intuitive data experiences, enable complex analysis through natural language, and even orchestrate conversational analytics agents as ‘tools’ for an orchestrator agent using Agent Development Kit.
How-tos
Integrate agentic workflows into your applications
The Conversational Analytics API lets developers embed natural-language query functionality in custom applications, internal tools, or workflows, all backed by trusted data access and scalable, reliable data modeling. It’s the same API that powers the out-of-the-box conversational experiences in Looker and BigQuery.
The Conversational Analytics API lets you build custom data experiences that provide data, chart, and text answers while leveraging Looker's trusted semantic model for accuracy or providing critical business and data context to agents in BigQuery. You can embed this functionality to create intuitive data experiences, enable complex analysis through natural language, and even orchestrate conversational analytics agents as ‘tools’ for an orchestrator agent using Agent Development Kit.
Agent development tools
Streamline how AI agents interact with your data
Google Cloud's agent development tools reduce the need for developers to build custom database connectors through ADK and MCP integration methods.
The MCP server for BigQuery allows an AI agent and MCP clients to interpret schemas and execute queries against BigQuery data while reducing the security or governance risks or latency associated with moving data into context windows.
For more flexibility and control, use the MCP Toolbox–an open-source server that centralizes the hosting and management of toolsets, decoupling agentic applications from direct database interaction. It’s also available with a variety of IDEs and developer tools including Gemini CLI and Antigravity, allowing you to securely connect your AI agents to services like AlloyDB, BigQuery, Spanner, Looker, and more.
Additionally, the BigQuery ADK integration toolset includes ready-to-use functions that enable agents to autonomously: explore data, understand schemas, run queries and forecasts, and get insights using natural language.
How-tos
Streamline how AI agents interact with your data
Google Cloud's agent development tools reduce the need for developers to build custom database connectors through ADK and MCP integration methods.
The MCP server for BigQuery allows an AI agent and MCP clients to interpret schemas and execute queries against BigQuery data while reducing the security or governance risks or latency associated with moving data into context windows.
For more flexibility and control, use the MCP Toolbox–an open-source server that centralizes the hosting and management of toolsets, decoupling agentic applications from direct database interaction. It’s also available with a variety of IDEs and developer tools including Gemini CLI and Antigravity, allowing you to securely connect your AI agents to services like AlloyDB, BigQuery, Spanner, Looker, and more.
Additionally, the BigQuery ADK integration toolset includes ready-to-use functions that enable agents to autonomously: explore data, understand schemas, run queries and forecasts, and get insights using natural language.
FAQ
What are the 4 types of data analytics?
AI agents enhance the four traditional pillars of data analysis:
- Descriptive: Explaining what happened (example summarizing a CSV)
- Diagnostic: Explaining why it happened (example, finding the root cause of a traffic drop)
- Predictive: Forecasting what will happen (example trend projection)
- Prescriptive: Suggesting what to do next (example, recommending budget shifts)
What are the 7 types of AI agents?
While definitions vary, AI agents are generally categorized by their complexity and behavior:
- Simple reflex agents
- Model-based reflex agents
- Goal-based agents
- Utility-based agents
- Learning agents
- Hierarchical agents
- Multi-agent systems (where multiple specialized agents collaborate)
How do AI agents automate the end-to-end data workflow?
Agents "close the loop" by handling the entire process: they ingest raw data, clean it, generate a plan for analysis, write the necessary scripts to process it, and then produce a final report or visualization without manual intervention.
How do I build an AI agent for data analysis?
Developers typically use frameworks like LangChain or Google Cloud's Agent Development Kit (ADK). The process involves connecting an LLM to a data source, providing it with "tools" (like a Python interpreter or SQL executor), and defining a system prompt that guides its reasoning.
