Skip to content
Theme
OnDevice Tools

SoftwareCloud & Services

Vodafone cuts data pipeline costs by 30 percent using Google Cloud agent

Vodafone Czech Republic has used Google Cloud's Data Engineering Agent to modernise BigQuery stored procedures, cutting pipeline execution times by half and reducing costs by 30 percent, according to the Google Cloud Blog.

Illustration of a data pipeline diagram showing cost reduction and faster execution times

Key takeaways

  • Vodafone Czech Republic cut data pipeline costs by 30 percent using Google Cloud's Data Engineering Agent.
  • Pipeline execution times were reduced by half and migration time fell by 90 percent compared with manual work.
  • The agent is available in BigQuery pipelines and Dataform and uses Gemini to automate workflows.
  • Vodafone plans to expand the agent to markets including Romania, Ireland, Germany and the United Kingdom.
In this article
  1. What Vodafone did
  2. Results and trade-offs
  3. How the agent works
  4. Vodafone's rollout plans
  5. Why it matters

What Vodafone did

Vodafone Czech Republic used Google Cloud's Data Engineering Agent to move legacy BigQuery stored procedures into modern Dataform pipelines. The agent, which uses Gemini to automate data engineering workflows, helped the company replace thousands of lines of sequential code with modular pipelines that run in parallel.

Results and trade-offs

The migration time fell by 90 percent compared with manual work, and the new pipelines run twice as fast as the old stored procedures. Caching intermediate tables using BigQuery materialized views led to a 30 percent reduction in costs. However, parallelising code initially increased costs because each thread performed full table scans instead of reusing cached base tables.

How the agent works

Illustration: AI, cloud, Google Cloud logo, Google logo and the labels Bigquery, Dataform, Google Cloud

The Data Engineering Agent takes natural language instructions and any provided code to build Dataform pipelines. It can understand dependencies between tables and cache intermediate tables using BigQuery materialized views. Dataform handles pipeline dependencies and execution logs, removing the need for manual metadata tables and spreadsheets.

Vodafone's rollout plans

Vodafone plans to scale the Data Engineering Agent across several markets, including the Czech Republic, Romania, Ireland, Germany and the United Kingdom. The company also wants to create a standardized Dataform template repository, integrate the agent into developer environments such as VS Code, and migrate legacy workloads such as SAS code, BTEQ scripts and MicroStrategy into Dataform pipelines.

Why it matters

The case shows how generative AI tools can reduce the toil of migrating legacy data systems. Vodafone engineers can now focus on higher-value tasks instead of repetitive migration work, while the company reduces technical debt across its data estate.

Sources

  1. How Vodafone is using the Data Engineering Agent to improve their data pipelinesGoogle Cloud Blog

Tags

Spotted a mistake? See our corrections policy. Read how News by OnDevice works in our editorial policy.