The next wave of data and AI is agentic, and the platforms running enterprise data are evolving fast.

Tomas Talius shares what's new in BigQuery and how the platform is being built for the agentic AI era, including the capabilities reshaping cost, performance, AI, and cross-cloud access.

Then Ofir Ehrlich joins to discuss a challenge most enterprises still face: critical data fragmented across business units and clouds, expensive to extract and hard to govern. He'll share where things are headed and what Eon and Google Cloud are doing together to change it.

A look at what's new, what's next, and what it means for the teams running BigQuery at scale.

Integrating AI into your data stack shouldn't require constant context-switching between notebooks and external APIs. This session is a technical walkthrough for data practitioners looking to leverage BigQuery AI directly where their data lives. We will focus on how to use the latest Generative AI functions to transform, enrich, and extract insights from your datasets using standard SQL.

Building a fast data platform is easy. Building one that is secure, scalable, and cost-efficient at scale is where most teams struggle.

In this session, we break down how to design modern, production-ready data platforms using Google BigQuery that don’t just work, but perform efficiently under real-world demands. We’ll explore practical strategies for query optimization, partitioning, and clustering to improve performance, while also implementing fine-grained security controls such as row-level and column-level access.

Beyond performance and security, we’ll address the often-overlooked dimension of cost—highlighting key trade-offs and design decisions that directly impact efficiency at scale.

You’ll walk away with actionable patterns, common pitfalls to avoid, and a clear blueprint to build data platforms that balance speed, security, and cost—ready for modern analytics and AI-driven workloads.

BigQuery makes it easy to scale analytics, and just as easy to scale your bill. In this session, Pedro Conde shares a practical playbook to make BigQuery spend predictable: the cost drivers that matter, the query and storage patterns that quietly inflate spend, and the guardrails teams use to reduce cost without sacrificing performance. You’ll also see a FinOps monitoring approach (dashboards + alerts) that helps teams spot anomalies early and drive accountability across projects. You’ll leave with concrete steps you can apply immediately in your own BigQuery environment.

Dataform is BigQuery's built-in framework for managing SQL transformations, but most teams either don't know it exists or underestimate how far it goes. In this hands-on session, Constantin walks through Dataform from two angles: the BigQuery Studio console experience for quick iteration, and the VS Code workflow for teams that want proper local development, Git integration, and code review. You'll see how to define sources, build SQLX models, manage dependencies, write data quality assertions, and schedule pipelines. Then, get an honest look at where Dataform excels and where it still has rough edges. Whether you're an analyst, data engineer, or analytics engineer, you'll leave with a working mental model and a clear sense of whether Dataform belongs in your stack.

Modern organizations rely on BigQuery as a mission-critical analytics platform, but data loss, corruption, or accidental changes can still happen. While BigQuery provides strong built-in durability and short-term recovery, it is not a complete backup solution.

In this session, we’ll walk through a practical “rollback playbook” for BigQuery - covering native capabilities like Time Travel and snapshots, their limitations, and how to design a robust enterprise-grade backup and recovery strategy. We’ll also explore real-world disaster scenarios and how to recover quickly while minimizing data loss and downtime.

In this closing panel session, three working practitioners will get into the realities of running BigQuery at scale. We'll cover cost surprises and where teams are getting the bill wrong, access management as BigQuery becomes more accessible to non-technical users and AI workloads, and where the platform is going next.