Salesforce is unveiling a comprehensive suite of products to strengthen its presence in the highly competitive AI industry. Known as AI Cloud, this suite encompasses a range of tools specifically designed to provide AI solutions that meet enterprise standards. With this strategic move, Salesforce aims to further enhance its product portfolio by incorporating AI capabilities. The launch of AI Cloud builds upon the company’s previous generative AI initiative, which was introduced in March and aimed to integrate generative AI across the entirety of the Salesforce platform.
Adam Caplan, SVP of emerging technology at Salesforce, said in a phone interview, “It’s really about bringing generative AI in a trusted fashion to the enterprise. We’re moving incredibly fast to leverage the history we have in AI and build this into into our stack in a trusted fashion.”
Salesforce’s AI Cloud is revolutionizing the hosting and delivery of AI models, specifically text-generating models, through a collaboration with esteemed partners such as Amazon Web Services, Anthropic, Cohere, and OpenAI. By leveraging the robust cloud infrastructure of Salesforce, these partners can effortlessly deploy and serve their AI models to customers. Salesforce’s AI research division has also developed first-party models that empower users with cutting-edge capabilities like code generation and business process automation.
One striking feature of AI Cloud is its flexibility, allowing customers to choose between utilizing the pre-existing models offered by Salesforce’s partners or bringing their custom-trained models onto the platform. This unique approach grants businesses the freedom to leverage their data infrastructure while harnessing the power and convenience of the AI Cloud.
Flexibility and Customizability in AI Cloud
Interestingly, this concept resembles Amazon’s recently launched Bedrock, which also offers a suite of models developed in-house by AWS alongside pre-trained models from startup partners. However, Salesforce’s approach emphasizes an ecosystem-based and open process, prioritizing the selection of the most suitable model for each use case. David Caplan, a representative from Salesforce, explained that their objective is to collaborate with the best models in the industry to provide optimal solutions.
In summary, Salesforce’s AI Cloud is poised to transform the AI landscape by offering a wide range of AI models, fostering collaboration with renowned partners, and providing customers the flexibility to bring their models. This strategic move solidifies Salesforce’s commitment to delivering cutting-edge AI capabilities to enterprises and highlights their dedication to an open and collaborative ecosystem.
Salesforce has developed a series of AI models within its AI Cloud platform that enhance various Salesforce products like Data Cloud, Tableau, Flow, and MuleSoft. These models consist of nine variants: Sales GPT, Service GPT, Marketing GPT, Commerce GPT, Slack GPT, Tableau GPT, Flow GPT, and Apex GPT.
Sales GPT enables the automated creation of personalized emails designed for sales purposes. Service GPT can generate service briefings, case summaries, and work orders using case data and customer history. Marketing GPT and Commerce GPT can create audience segments for targeted marketing and tailor product descriptions based on customer data. Additionally, they can provide recommendations on how to increase average order value.
Einstein Trust Layer: Enhancing Data Privacy and Compliance
The other models, namely Slack GPT, Tableau GPT, Flow GPT, and Apex GPT have more specialized functionalities. Slack GPT and Flow GPT empower users to create no-code workflows integrated with AI actions within Slack or Flow. Tableau GPT can generate visualizations and data insights based on natural language prompts. Apex GPT, on the other hand, focuses on scanning for code vulnerabilities and suggesting inline code for Apex, which is Salesforce’s proprietary programming language.
Several models are already in use, including Slack GPT, Commerce GPT, Sales GPT, and Service GPT. The remaining models, except Flow GPT, which is scheduled for release in October, are expected to be available as early as this month.
AI Cloud offers a distinctive feature called the Einstein Trust Layer, an Artificial Intelligence moderation and redaction service introduced by Salesforce. This service addresses concerns about retaining sensitive data by text-generating models, such as customer purchase orders and phone numbers. Like Nvidia’s NeMo Guardrails, the Einstein Trust Layer aims to ensure that companies with stringent compliance and governance requirements can safely utilize generative AI tools.
This offering is particularly relevant considering the current landscape. Many companies, including Amazon, Goldman Sachs, and Verizon, have implemented bans or restrictions on using generative AI tools like ChatGPT due to privacy risks. By providing the Einstein Trust Layer, Salesforce aims to enable organizations with strict compliance standards to leverage productive AI capabilities while maintaining data privacy and security.