Salesforce recognizes that customers could have data stored in external systems, as well as custom AI models. The new partnership with Google Cloud allows organizations to bring data and large language models (LLMs) from Google, to support what happens within Salesforce Data Cloud.
Data Cloud is Salesforce’s CDP (customer data platform) offering – but with a twist on the typical CDP. Data Cloud is at the foundation that connects different ‘clouds’ across the platform. When customers directly purchase/license Data Cloud, they gain an even stronger setup with Data Cloud as a personalization and data unification platform. Most recently, Data Cloud has had a big role in the Marketing GPT and Commerce GPT innovations.
The partnership with Google Cloud is two-fold – reduce the complexity of moving data between platforms with zero-copy ETL between BigQuery and Data Cloud, and ‘bring your own’ AI model from Vertex AI. Both of these connections reduce the cost, risk, and complexity of syncing data across platforms.
Data Cloud + Vertex AI
Build AI models in Vertex AI, then bring those models into Salesforce to be used alongside Salesforce data. Vertex AI enables analysts to manage machine learning (ML) projects throughout the entire development life cycle, including training, testing, monitoring, deploying, and governing models at scale.
On the flip-side, zero-copy data access will streamline AI model training. Without replicating your CRM data in Google Cloud, you can still leverage CRM data to train AI models, therefore improve the accuracy and efficiency of AI predictions and insights.
Use cases for how custom AI models from Google Cloud can be ‘unleashed’ within the Salesforce platform (across the entire Customer 360) range from predicting buying behavior or churn likelihood, to personalized recommendations, and enhancing chatbots:
Availability: In pilot in July 2023, generally available in October 2023.
Data Cloud + BigQuery
A secure link between Data Cloud and the data outside of the Salesforce platform – to have it show up in Salesforce as if it is stored there (known as zero-copy or zero-ETL (Extract, Transform, Load)). This reduces the complexity of your data architecture, the storage costs of not duplicating records – all while maintaining governance and trust between platforms.
You can query first-party (known) CRM data, and other data stored in the Google Cloud Platform, directly and use within Salesforce Data Cloud.
Availability: In pilot in October 2023, generally available in February 2024.