Supporting generative AI development with our data cloud partners

Today, our data and analytics partners are rolling out new updates and integrations with Google Cloud to help customers significantly reduce the time it takes to go from managing data to building new AI tools and workflows. These partner solutions represent a significant step forward in Google Cloud’s mission to offer the industry’s most open data cloud ecosystem—a major asset for customers building generative AI. 

By utilizing Google Cloud services like BigQuery, our powerful foundation models in Vertex AI, and our trusted infrastructure, partners can help customers improve some of their most common (and complex) use cases for data. Data and analytics companies are then providing more tooling to help businesses address industry-specific needs, develop bespoke generative AI models, and apply AI to help users more efficiently explore their data.

Supporting industry-specific use cases for generative AI

Businesses have an opportunity to use generative AI to address common, industry-specific challenges, such as helping a healthcare provider instantly generate patient notes for clinician review, or enabling a loan officer to instantly generate a summary of open applications. Data underpins all of these industry scenarios, and now our partners are launching several new solutions to address them.

• Confluent, which provides a widely-adopted data streaming platform, will use Google Cloud’s generative AI to launch new solutions for retail and financial services customers, improving business insights and operational efficiencies. For example, generative AI can help an inventory manager better predict supply shortages based on local demand trends to mitigate “out of stock” items, while a banking security expert can apply it to improve fraud detection based on more accurate risk models and alerts.

• DataRobot, a platform for building AI solutions, will use Vertex AI to help clinicians better understand and make decisions with their data. By querying a patient’s name and symptoms, a new solution will run a model that uses BigQuery for analytics and can reference personalized medical history stored securely on AlloyDB to support a clinician in determining next steps in care. The solution can give healthcare organizations full control over their data, with preview access available to customers later this year.

• MongoDB, a leading developer data company, is working with Exafluence to develop a platform called Exf ChemXpert for the chemical industry to help researchers and chemists plan for synthesis and predict forward reaction, reaction completion, and retrosynthetic routes. The platform will use AI and data mining techniques with MongoDB Atlas Vector Search and foundation models from Google Cloud to discover new molecules. Exf ChemXpert will also include configurable components for a wide variety of applications in the chemical industry, such as property prediction to guide the design of new molecules, chemical reaction optimization to make developing molecules more environmentally friendly, and novel drug discovery in the pharmaceutical industry. 

Helping customers build custom generative AI models and applications

Partners like Dataiku, Redis, SingleStore, and Starburst are all making it easier for customers to train AI models and build new generative AI applications using data stored within their platforms. Now, each of these partners is working to help customers apply powerful AI models on top of their relevant data.

• Dataiku, an AI and data science platform, will integrate with Vertex AI to bring the PaLM 2 foundation model to its Prompt Studios interface to improve how engineers design, test, and operationalize generative AI prompts. With PaLM 2, users will be able to easily deploy a powerful model directly within Dataiku workflows, which enables enterprises to more quickly build and scale their generative AI applications. 

• Redis, which delivers an enterprise-grade data platform, will help customers build custom generative AI applications together with Vertex AI. By utilizing contextual data stored in Redis as a vector database, and Google Cloud's powerful foundation models, customers will be able to create unique generative AI applications like virtual shopping assistants, automated customer support services, and more. 

• SingleStore, a cloud-first database system, will utilize Vertex AI to help customers build bespoke generative AI applications based on relevant data stored across its systems. This will allow SingleStore users to more easily identify and contextualize real-time business insights through experiences like a chatbot, which will apply Google Cloud’s natural language processing to provide more contextual, accurate responses. SingleStore will launch these capabilities later this year.

• Starburst, the fast-growing data lake analytics platform, will integrate with Vertex AI to enable customers to build and train generative AI models from data stored across multiple cloud providers and on-premises environments. Starburst’s federation engine unites this data in a single environment, where customers will be able to run advanced analytics workloads and improve analysis with Google Cloud’s foundation models. 

Exploring large datasets with generative AI

Generative AI can further democratize data by letting more people explore large datasets with accuracy and confidence. With Vertex AI, many of our partners are utilizing Google Cloud’s LLMs and other foundation models in Model Garden to enhance how users search and analyze data stored across a wide array of systems. 

• Datastax, which provides a scalable vector database, will integrate with Vertex AI to help customers build secure, production-ready AI applications. Through a new extension in DataStax's Astra DB, developers will be able to easily combine data stored in Astra DB with Google Cloud’s foundation models to enable accurate, consistent, and contextual responses to search queries with natural language understanding. The new capabilities will launch this Fall. 

• Elastic, a leading platform for search-powered solutions, will integrate Vertex AI with the Elasticsearch Relevance Engine. By combining enterprise data with Google Cloud’s foundation models, Elastic users can have a secure deployment that can generate relevant, factual answers to users’ complex questions. Elastic will launch these capabilities later this year.

• Neo4j, a leading graph database provider, will integrate with Vertex AI to add vector search to its AuraDB database offering. This capability can improve the search experience by providing fast performance and contextual and accurate results—for example, a consumer searching for umbrellas on a retail customer’s website could also see results for rain boots and jackets. AuraDB customers can start using these features next week.

We want to provide customers with the industry’s most open, flexible cloud platform, and that is echoed in our work in generative AI. New integrations with data and analytics companies will be invaluable in helping to extend the power of generative AI to more customers, and we believe they can dramatically improve how businesses use their data moving forward.