Google expands BigQuery with Gemini, brings vector support to cloud databases

Key Points:

  • Google expands capabilities of its database and analytics platforms with generative AI
  • Vector search support now available across all Google cloud databases
  • Enhancements include Gemini Pro models for advanced analytics in BigQuery


In 2024, Google has been actively expanding its database and analytics platforms to leverage generative AI capabilities. Notably, Google has integrated its Gemini large language models into its BigQuery analytics service, introducing new functions for AI data preparation and retrieval augmented generation (RAG). Additionally, Google is broadening its database offerings by introducing vector search support across all cloud databases.


Andi Gutmans, GM and VP for Databases at Google Cloud, emphasized the importance of incorporating vector indexing and search as a fundamental component in databases to harness the full potential of AI applications in the enterprise. Google has already incorporated vector support in some of its databases, with AlloyDB database now generally available with enhanced vector and AI capabilities. Moreover, Google is expanding vector support to include other databases like Memorystore for Redis, CloudSQL, Spanner, Firestore, and Bigtable.


Developing vector support across all Google databases entails substantial engineering effort, according to Gutmans, particularly in customizing solutions for different database structures. Google’s expertise in building scalable vector-capable indexes is a distinguishing factor, accumulated over years of internal usage in ad and search business units. Gutmans highlighted Google’s advantage in this area, given its extensive experience in managing vector search capabilities at a large scale.


On the analytics front, Google is enhancing BigQuery with support for Gemini Pro models, enabling advanced functionalities like summarization, sentiment extraction, classification, and data enrichment for structured and unstructured data. Gerrit Kazmaier, GM and VP for Data Analytics, described these enhancements as unlocking new analytical possibilities by enabling comprehensive analysis of unstructured data alongside structured data within BigQuery.


The move to infuse AI capabilities into database and analytics services aligns with Google’s commitment to empowering developers and organizations to leverage the benefits of generative AI effectively. By incorporating advanced technologies like vector search and Gemini models, Google aims to provide users with enhanced tools for data analysis and retrieval, catering to the evolving needs of enterprises in leveraging AI-driven insights.



Prompt Engineering Guides



©2024 The Horizon