BigQuery autonomous data to AI platform uses AI to automate data analysis, transformation, and insight generation. It simplifies complex workflows with built-in intelligence and natural language interaction.
The fast-paced world need a real-time data activation flywheel in addition to data access. Artificial Intelligence that is integrated directly into the data environment and collaborates with intelligent agents is a new reality that is beginning to take shape. These agents serve as catalysts, opening doors for everyone and facilitating the self-directed, instantaneous action that is essential for success. This flywheel is powered by Google’s Data & AI Cloud, which applies AI to data for continuous, real-time data activation. Because of this emphasis, BigQuery is drawing five times as many organizations as the two top cloud providers that just provide data science and data warehousing platforms.
Leading corporations include, for example:
- By fine-tuning the Gemini model with BigQuery, Radisson Hotel Group increased campaign productivity by 50% and revenue by more than 20%.
- Gordon Food Service created a scalable, contemporary data architecture with an AI-ready basis by integrating more than 170 data sources using BigQuery. In addition to lowering expenses and increasing market share, this enhanced real-time reaction to crucial business demands, made full analytics possible, significantly increased client use of their ordering systems, and gave their staff immediate insights.
- By merging disparate technologies, such as Databricks, into a single BigQuery platform, J.B. Hunt is revolutionising logistics for shippers and carriers.
- By providing workers with safe access to LLMs to respond to queries based on both structured and unstructured data, General Mills saves more than $100 million through the use of BigQuery and Vertex AI.
Together with unified, trustworthy, and conversational BI platform with Looker, Google Cloud is launching a number of new advancements with their autonomous data to AI platform powered by BigQuery:
- Agents with specific expertise for each user: The job of data scientists, data engineers, analysts, and business users will be made easier and faster by new assistive and agentic experiences that are based on your trusted data and accessible through BigQuery and Looker.
- Accelerating data science and advanced analytics: In addition to seamless connection with real-time and open-source technologies, It is improving data science processes in BigQuery with new AI-assisted notebooks and revealing new insights with BigQuery AI Query Engine.
- Autonomous data foundation: BigQuery’s new autonomous features, which include native support for unstructured data processing and open data formats like Iceberg, let it to acquire, handle, and orchestrate any forms of data.
Let’s examine each of these changes in more detail.
Agents with specialized knowledge for each user
It think everyone should have access to AI. After making AI-powered helpful experiences widely accessible in BigQuery and Looker, Google Cloud has now extended reach to include specialized agents that best suit the requirements of all data jobs, such as:
- Integrated with BigQuery pipelines, data engineering agent capabilities assist the creation of data pipelines, carry out data preparation tasks such as data transformation and enrichment, preserve data quality through anomaly detection, and automate metadata creation. These agents replace time-consuming and repetitive activities and allow trustworthy data, increasing the productivity of your data teams. Traditionally, data engineers spend endless hours cleaning, processing, and verifying data.
- Every step of model construction is made possible by the data science agent, which is integrated into Google’s Colab notebook. It facilitates scalable training, intelligent model selection, automated feature engineering, and quicker iteration. Instead of battling with data and infrastructure, this agent enables data science teams to concentrate on creating sophisticated data science procedures.
- All users may engage with data using natural language with Looker conversational analytics (preview). Expanded capabilities created in collaboration with DeepMind enable all users to comprehend the behaviour of the agent and easily resolve misunderstandings by doing sophisticated analysis and clearly explaining its reasoning. Additionally, accuracy is increased by up to two thirds because to Looker’s semantic layer. When users use business terminology like “revenue” or “segments,” the agent understands exactly what you mean and is able to compute metrics in real-time, guaranteeing that the results are reliable, accurate, and pertinent. It is also introducing a conversational analytics API so that developers may create and incorporate conversational analytics into workflows and apps.
Google Cloud also unveiling the BigQuery knowledge engine to fuel intelligence across assistive and agentic experiences in the BigQuery autonomous data to AI platform. It makes use of Gemini’s capabilities to examine table descriptions, query histories, and schema linkages in order to model data associations, suggest business glossary words, and create metadata instantly. This knowledge engine, which grounds AI and agents in business context, serves as the basis for AI-powered experiences, such as semantic search across BigQuery and AI-powered data insights.
All customers may now use all of Gemini-powered agentic and assistive experiences in BigQuery and Looker inside the current price model tiers without the need for add-ons!
Accelerating advanced analytics and data science
By allowing new AI-driven data science experiences and new engines to handle complicated data and enable sophisticated analytics in real-time, BigQuery autonomous data to AI platform is radically altering the way data scientists and analysts operate.
First, they are using AI to enhance the BigQuery notebook experience. It is adding intelligent SQL cells that can combine data sources right within your notebook, understand the context of your data, and provide intelligent suggestions as you write code. Additionally, in order to facilitate data exploration and facilitate collaboration with peers, it are incorporating native exploratory analysis and visualization tools. Additionally, data scientists may plan analyses to run and frequently update insights. Additionally, Google Cloud is offering the capability to create dynamic, user-friendly, interactive data apps that are driven by your laptop in order to disseminate insights more widely within the organization.
It is also introducing the BigQuery AI query engine to provide sophisticated, AI-driven analytics, building on this improved notebook environment. With the help of this engine, data scientists can handle both structured and unstructured data with ease and add real-world context, going beyond just retrieving structured data. Traditional SQL and Gemini are co-processed by the BigQuery AI query engine, which adds runtime access to linguistic comprehension, reasoning skills, and real-world knowledge. and the unstructured photos are processed by their new engine, which then compares them to your product catalogue. A wide number of use cases are supported by this engine, including as creating richer features for models, carrying out complex segmentation, and gaining previously unattainable insights.
Additionally, it give consumers access to the finest open-source ecosystem, optimised for the cloud. Google Cloud for Apache Kafka powers serverless execution of Apache Spark workloads within BigQuery by enabling real-time data pipelines for event sourcing, model scoring, communications, and real-time analytics. In the last year, customer usage of their serverless Spark capabilities has almost quadrupled, and Google Cloud has improved this engine to process data 2.7 times quicker than it did the year before.
Whether it’s SQL, Spark, or the semantic power of foundation models, BigQuery enables data scientists to use the tools they require on Google’s serverless and scalable architecture, facilitating quicker innovation without the difficulties associated with traditional infrastructure.
An independent data foundation throughout the duration of the data lifecycle
Its sophisticated analytics engines and specialized agents are supported by an independent data foundation built for the complexity of contemporary data. By elevating unstructured data to the status of first-class citizens within BigQuery, it is radically altering the environment. This is made possible by the platform’s new features, which include orchestration for a variety of data workloads, autonomous and invisible governance, and a dedication to flexibility through open formats, guaranteeing that your data is always prepared for any data science or artificial intelligence issue. And it accomplish all of this while providing the best possible pricing performance and reducing operational overhead.
For many organizations, the potential of their unstructured data represents their largest unrealized opportunity. Unique ideas contained in text, audio, video, and photos are frequently difficult to extract, underutilized, and usually found in siloed systems, even if structured data provides analytical paths. BigQuery immediately addresses this issue by elevating unstructured data to the status of a first-class citizen using multimodal tables (preview), which enables you to combine structured data and rich, complex data types for unified querying and storage.
Google Cloud’s improved BigQuery governance gives data stewards and professionals a single, unified view to manage discovery, classification, curation, quality, usage, and sharing, including automated cataloguing and metadata generation, in order to efficiently manage this extensive data estate. Furthermore, BigQuery continuous queries allow for immediate analysis and action on streaming data using SQL, independent of its original format, to guarantee timely insights from all your data streams.
Adoption is being driven by sophisticated support for both structured and unstructured multimodal data; year over year, customers’ use of Google’s AI models in BigQuery for multimodal analysis has increased by 16 times. Comparing their combined data and AI strategy to competing standalone data warehouse and AI systems, it find that BigQuery and Vertex AI are 8–16 times more cost-effective.
Google Cloud still steadfast in it commitment to an open ecology. You can link your Iceberg data to SQL, Spark, AI, and third-party engines in an open and interoperable way using BigQuery tables for Apache Iceberg, which offer the performance and integrated tools of BigQuery together with the freedom of an open data lakehouse. In addition to high-performance streaming, auto-AI-generated insights, almost limitless serverless scaling, and superior governance, this service offers adaptive and autonomous table management. Their managed solution offers fail-safe features and centralized fine-grained access control management through interaction with cloud storage.
Lastly, the AI platform’s autonomous data optimizes itself. With its comprehensive workload management capabilities, it scales resources, manages workloads, and helps guarantee their cost-effectiveness. Additionally, it has made purchasing easier with the new BigQuery spend commit, which unifies expenditure throughout BigQuery platform and offers flexibility in moving spend across streaming, governance, data processing engines, and more.
Use BigQuery data transfer solution to get started on your data and AI journey. Google Cloud eager to find out how you are leveraging data to innovate.