Salesforce, the developer of the CRM system of the same name that is in demand today, has announced the release of a novelty – a corporate set of AI Cloud services. The kit is presented with tools through which you can create and deploy AI models for enterprises in accordance with their individual needs and wishes. Famous Salesforce partners such as Amazon Web Services, OpenAI, Anthropic, Cohere, and others already use the AI Cloud to host and support AI models. If necessary, the data of the specially trained AI model hosted here can be stored separately, in your infrastructure.
Salesforce models located in the AI Cloud generate program codes, automate routine business processes, and provide useful recommendations, significantly expanding the capabilities of the company's previously released products:
- Sales GPT (writes commercial offers and texts for mailing lists in seconds in accordance with the rules of personalized marketing).
- Service GPT (after analyzing client data, it compiles briefings to improve the quality of service and also makes summaries of orders and cases).
- Commerce GPT and Marketing GPT (based on the results obtained after analyzing client data, optimally segment the audience to achieve maximum targeting efficiency, personalize the shopping experience, provide customers with individual offers, develop effective marketing strategies, and also provide valuable recommendations, for example, what to do to increase the average check).
- Flow GPT and Slack GPT (allow you to create workflows using AI, eliminating the need to write program code yourself).
- Tableau GPT (creates visualizations – charts and interactive dashboards – based on the results of data analysis and human prompts).
- Apex GPT (checks the program code for errors that lead to vulnerabilities and also offers the user inline code for the Apex language with Java-like syntax on the Salesforce CRM platform).
The Sales GPT, Service GPT, Commerce GPT, and Slack GPT models can be used today. The company will launch the rest by the end of this month. The only exception is Flow GPT, which will appear no earlier than October 2023.
An attractive feature of the AI Cloud toolkit is the presence of the Einstein Trust Layer moderator service in its composition. It prevents generative AI models from storing sensitive user information (for example, phone numbers or order details). Einstein is a checkpoint between an online service or mobile application and a generative AI model. It "sifts" the data being sent, looking for confidential ones among them, and if they are found, it immediately deletes them even before they enter the GPT. In addition, the Einstein Trust Layer filters out "toxins", thus counteracting discrimination (gender, ethnicity, sexual orientation, and its other forms). It removes the text fragments containing them not only in the hints formulated by the person but also in the answers given by the generator.
The company Salesforce reported that those who wish will be able to use the Einstein Trust Layer service towards the end of this month. The AI Cloud Enterprise Toolkit is scheduled to be released before the end of 2023.