To bring models to the enterprise, Salesforce unveils AI Cloud

Salesforce recently unveiled a new line of products designed to strengthen its position in the fiercely competitive AI market.

Named AI Cloud, the suite consists of tools designed to deliver “enterprise ready” AI.  The suite of tools is Salesforce’s latest cross-disciplinary attempt to enhance its product offering with AI capabilities. In several ways, it’s a continuation of the generative AI initiative that the company launched in March 2023, which intends to incorporate generative AI across Salesforce’s entire platform.

Over a phone call, Senior Vice President of Emerging Technology at Salesforce, Adam Caplan, told reporters that it was really about bringing generative AI in a trusted fashion to the business, and Salesforce was moving incredibly fast to leverage the history the firm has in AI and build it into their stack in a trusted fashion.

Salesforce’s AI Cloud hosts and serves AI models from a wide range of partners, including Anthropic, Amazon Web Services, OpenAI, and Cohere, on Salesforce’s cloud infrastructure. Salesforce’s AI research division provides first-party models that can power capabilities such as code generation and business process automation. Customers can still use Salesforce's AI platform while storing data on their own cloud infrastructure by bringing a specially trained model to it.

The idea is similar to Amazon's recently released Bedrock, which offers a family of models that have been trained internally by AWS as well as models that have already been trained by startup partners.

New features in Salesforce's flagship products, such as Data Cloud, Tableau, Flow, and MuleSoft, are made possible by models created by Salesforce and included in AI Cloud. Commerce GPT, Slack GPT, Sales GPT, Service GPT, Marketing GPT, Tableau GPT, Flow GPT, and Apex GPT are the nine models in total.