BigQuery's AI Tools Make Writing Pipelines Faster. Maia Now Governs These Pipelines.

Maia Foundation, the execution layer of Matillion's AI Data Automation platform, runs on Google BigQuery as of today. Net-new BigQuery environments can onboard onto Maia Foundation directly. Existing Matillion ETL for BigQuery customers keep running as they are, with a guided migration path when they choose to move.

Why "AI in BigQuery" Still Leaves Your Team Building Pipelines

Nearly every AI feature shipping into the BigQuery ecosystem this year speeds up engineers writing SQL, models, and prompts. Gemini Code Assist in BigQuery accelerates authorship. dbt's AI features accelerate authorship. Coalesce's Copilot, Prophecy's agentic copilot, the horizontal coding tools sitting in your team's IDE, all of them accelerate authorship.

That answer assumes the bottleneck is how fast pipelines get written. It isn't. The pitch is throughput: same engineer, same role, more lines per hour.

Writing the pipeline isn't the only tricky part of the data engineering lifecycle. The focus on pipeline generation undercuts the complexity of the work around it: the review, the data contracts, the quality checks, the governance sign-off, the promotion window, the work of earning downstream trust. Speed up authorship alone and you have a faster first draft feeding the same queue. Gene Amdahl named this in 1967: the total speedup of a system is capped by the part you didn't speed up. For data teams, the part nobody automated is the work around the pipeline, and that is most of the work.

Maia automates that surrounding work, not just the authoring. Maia builds, governs, and maintains the pipelines; the engineer reviews and approves. Different role, different operating model, different shape of the team next quarter.

What Maia Does on BigQuery

Maia is built around three layers, each with a single job:

• Maia Team authors pipelines. Autonomous agents take business intent and the structure of your data and produce orchestration and transformation logic that runs against BigQuery.
• Maia Context Engine keeps the work governed. It maintains a current model of your schemas, lineage, and rules so the automation stays inside the lines your governance team has drawn.
• Maia Foundation executes. Transformations compile to SQL and run inside BigQuery. The cost and performance of BigQuery are preserved. Lineage and observability are emitted as the pipelines run, not bolted on after.

One concrete example. When an upstream source renames a column, Maia detects the change, rebinds the downstream transforms in line with the governance rules in the Context Engine, regenerates the lineage entry, and commits the change to Git. An engineer reviews the change and approves it. The engineer did not write any of it. That single behavior repeated across hundreds of pipelines is the operating model change.

Heritage: Matillion Has Run BigQuery in Production for Years

Maia for BigQuery is not Matillion's first BigQuery launch. Matillion has been running BigQuery pipelines in enterprise production for years, feeding analytics that moved off legacy on-prem systems and onto BigQuery.

Matillion has been the ELT layer for BigQuery in enterprise production for years. Maia for BigQuery is the next platform underneath that practice, not the start of it.

Available Today on Google BigQuery

From this release, BigQuery customers can:

• Run Maia Foundation with core capabilities, including pushdown ELT into BigQuery, enterprise data connectivity, pipeline orchestration and scheduling, Git for DataOps, pipeline monitoring and observability.
• Move existing Matillion ETL for BigQuery projects across through a guided, engineering-supported migration path.
• Deploy via hybrid agent (data plane in your own environment) or full SaaS, depending on how the team needs to operate.
• Onboard directly through the Matillion Hub.

Starting on Maia Foundation Directly

Any BigQuery environment can onboard onto Maia Foundation directly, with the same execution layer that supports Maia on Snowflake, Databricks, and Redshift. Authorship runs through Maia Team's agents from day one. Governance is constructed as the pipelines are built, not added afterwards. This is the fastest way to experience Maia yourself, natively on BigQuery.

Migrating from Matillion ETL for BigQuery

Your existing Matillion ETL for BigQuery environment is supported and keeps running. When you choose to move, migration is a guided process: we start with a diagnostic of your specific project, then convert it onto Maia Foundation with our engineering team working alongside yours. You keep the pushdown-into-BigQuery model you already work with, and you gain Maia Team's authorship, the Context Engine's continuous view of your environment, and Mission Control's central operational plane. We are deliberately running a small number of migrations at a time to get each one right. The next step is a conversation with your account team about your environment.