Google Advances Cloud Data Platforms and Analytics Services with GenAI

The emergence of generative artificial intelligence (GenAI) has significant implications at all levels of the technology stack, not least analytics and data products, which serve to support the development, training and deployment of GenAI models, and also stand to benefit from the advances in automation enabled by GenAI. The intersection of analytics and data and GenAI was a significant focus of the recent Google Cloud Next ’24 event. My colleague David Menninger has already outlined the key announcements from the event as they relate to Google Cloud’s Vertex AI development platform and the Gemini multimodal large language models (LLMs), as well as Gemini-based GenAI assistants. In this perspective, I’ll take a closer look at Google Cloud’s progress as it relates to the company’s various data platforms, as well as the Looker analytics platform.

Google Cloud has a broad portfolio of data and analytics offerings that spans data storage and processing, data management and data governance, as well as analytics and machine learning (ML). The company has sought to differentiate itself from rival cloud providers in recent years by emphasizing its multi-cloud and hybrid architecture credentials, as well as its expertise in relation to data, analytics and AI. That emphasis was reiterated at Google Cloud Next ’24, with the company highlighting the BigQuery analytic data platform, the Looker analytics environment and the AlloyDB for PostgreSQL operational data platform. All three are available on multiple cloud services and were the beneficiaries of new digital assistant capabilities based on Google’s Gemini LLMs, providing new natural language interfaces for analyzing and managing data. Although only 41% of participants in ISG’s AI Buyer Behavior Study are piloting or in production with natural language queries, 88% have seen many or some positive outcomes. Meanwhile 39% of participants are piloting or in production with natural language interpretation of data, of which 87% have seen many or some positive outcomes.

Google BigQuery is a distributed serverless analytic data platform environment for processing and analyzing large volumes of data. It is now being positioned by Google as a unified data platform that provides multiple data engines — including SQL, Spark and Python — to process data in multiple formats, across multiple locations, for multiple use cases (including business intelligence (BI), as well as AI). In addition to being available on Google Cloud, BigQuery is also available on Amazon Web Services as well as Microsoft Azure via BigQuery Omni, while the BigLake storage engine enables enterprises to work with structured and unstructured data in open table formats, including Apache Iceberg, Delta and Hudi. Among the announcements at this year’s Google Cloud Next event was the preview release of Apache Kafka for BigQuery and the incorporation of enhanced search capabilities, courtesy of Google Dataplex’s unified metadata catalog.

Also new is direct access from BigQuery to Google’s Vertex AI development platform for AI, enabling vector search and retrieval augmented generation to ground the output of GenAI models with enterprise data, as well as the ability to fine-tune models in Vertex AI from BigQuery. In addition to using BigQuery to deliver and improve GenAI, enterprises can now also use GenAI to improve the use of BigQuery thanks to the introduction of Gemini in BigQuery and BigQuery data canvas. Gemini in BigQuery provides GenAI-based assistance for BigQuery users, including augmented data preparation, semantic search-based data exploration, the conversion of natural language queries to SQL or Python code, and recommendations to improve query performance. BigQuery data canvas is a natural language interface for data exploration, curation, preparation, analysis and visualization. Google also announced the general availability of the BigQuery Studio collaborative analytics workspace, which provides a single environment for analyzing data in BigQuery using SQL, Python, Spark or natural language queries.

Google Cloud also introduced Gemini in Looker and Gemini in Databases. Gemini in Looker provides an interface for natural language conversational analytics using the Looker analytics environment that is grounded by business data, with definitions provided by the LookML semantic modeling language. I recently noted the increased importance of semantic data modeling to standardize metrics and definitions as more business users begin to interact with and analyze enterprise data using GenAI interfaces. Gemini in Looker includes LookML Assistant, which automatically generates LookML code based on natural language requests. Gemini in Databases provides GenAI functionality to provide assistance for database administrators, including Database Studio for natural language SQL generation, Database Insights to optimize performance based on AI-powered recommendations, Database Center for managing an enterprise’s complete database fleet, and Database Migration Service for automated conversion of code and database schema from Oracle PL/SQL to Google Cloud SQL or AlloyDB for PostgreSQL. First announced in 2022, the fully-managed AlloyDB for PostgreSQL service has quickly become front and center of Google Cloud’s portfolio of operational database services. Available for deployment on premises and on other clouds via AlloyDB Omni, and with support for the development of GenAI applications via AlloyDB AI, the database service has been updated with a new ScaNN index to boost search performance, as well as native support for natural language queries.

Google Cloud is by no means unique in providing digital assistant capabilities based on GenAI for data and analytics. Indeed, I assert that through 2026, analytics and data software providers will prioritize the delivery of automated code and query generation and conversion capabilities based on GenAI. However, the company is ahead of several of its rivals in terms of the breadth of functionality on offer and the range of products for which AI assistants are available or in preview. Enterprises evaluating databases to support new application development projects, both on premises and in the cloud, should consider Google Cloud BigQuery and AlloyDB for PostgreSQL, as well as Google Cloud’s associated analytics and GenAI-based services.

Regards,

Matt Aslett