Google Cloud announces a series of new AI features aimed at deploying agents within data environments.
The agents are specialized software applications that can perform data tasks independently and are positioned as a complement to existing workflows. According to Google, the introduction marks a shift toward more autonomous collaboration between humans and machines in data processing.
For data engineers, a Data Engineering Agent will be available in BigQuery. This generates data pipelines based on natural language. This allows operations such as data loads, transformations, and quality checks to be largely automated. A Data Science Agent will also be available in the same environment, which will independently perform tasks such as data cleaning, feature engineering, and model training. The user remains involved through feedback and customization options.
Expansion of Conversational Analytics Agent
For business users, the Conversational Analytics Agent, previously launched, is being expanded with a Code Interpreter. This can convert questions in natural language into Python code and present results in the form of visualizations with textual explanations. This should enable more advanced analytics for users without programming knowledge.
In addition to introducing these agents, Google is also making resources available for developers who want to build their own agents or expand existing applications. APIs and a development kit will be available under the name Gemini Data Agents. These will allow companies to link their own agents to Google’s platform or incorporate conversation-driven functionality into their applications.
Underlying these announcements are changes to the data structure itself. For example, Google is introducing a column engine in Spanner, designed to significantly speed up analytical queries on transactional data. BigQuery will also gain support for vector search and automatic data embedding, which is essential for applications such as retrieval-augmented generation. A similar step is being taken in AlloyDB with optimizations for vector search on live data.
An AI Query Engine is also being added to BigQuery. This makes it possible to ask questions involving semantics or interpretation, such as analyzing customer feedback on tone or emotion, within an SQL query.