New Vector Database in Salesforce Data Cloud Will Power AI, Analytics, and Automation Using LLMs with Business Data for Use Across the Einstein 1 Platform

Salesforce announced significant updates to its Einstein 1 Platform, adding the Data Cloud Vector Database and Einstein Copilot Search.

Accurate and relevant generative AI prompts require grounding in the most comprehensive set of enterprise data. Until now, this has required expensive and labor-intensive model fine-tuning. Data Cloud Vector Database will solve this challenge by making it quick and easy to bring unified business data into any AI prompt so customers can deploy trusted, relevant generative AI across all Salesforce applications without having to fine-tune an off-the-shelf large language model (LLM).

Data Cloud Vector Database – built into the Einstein 1 Platform – enables AI, automation, and analytics for improved decision-making and customer insights across all Salesforce CRM applications. Data Cloud will also power Einstein Copilot Search — announced today — enhancing Einstein Copilot, Salesforce’s generative AI assistant, with AI search capabilities that use all business data to deliver more precise information, conveniently in the flow of work.

New capabilities:

Data Cloud Vector Database

Data Cloud Vector Database will remove the need to fine-tune LLMs by seamlessly using all business data to enrich AI prompts, allowing customers to use a variety of data types across their business applications and workflows. This increases business value and ROI by unifying unstructured data, including PDFs, emails, documents, and transcripts, with structured data, including purchase history, customer support cases, and product inventory, to power AI, automation, and analytics across every Salesforce application.

Also Read: Varonis Announces Data Security Platform on Salesforce AppExchange

For example, customer service leaders will enhance efficiency and customer satisfaction by utilizing a platform that proactively presents relevant knowledge articles to service agents the moment a case is created. This allows for quick identification of similar cases and the integration of automation, thereby reducing case resolution time and improving the overall customer experience.

Einstein Copilot Search

Einstein Copilot, available in February, will include enhanced AI search capabilities that interpret and respond to complex queries from users by tapping into diverse data sources, including unstructured data. Einstein Copilot Search will enhance Einstein Copilot, providing sales, customer service, marketing, commerce, and IT teams with an AI assistant capable of solving problems and generating content by accessing real-time unstructured and structured business data. Customers will benefit from an AI assistant that understands and addresses complex queries by accessing insights and knowledge previously unattainable with foundational LLMs due to limitations in their training data.

For example, in customer service, Einstein Copilot Search will link a customer’s concerns from unstructured emails and phone call transcripts to their structured support ticket history. This provides service representatives with a detailed understanding of customer issues and their historical context and AI-generated, data-backed resolution suggestions. And, the new integration of source citations enhance the customer service team’s confidence in the AI-generated insights.

Why it matters: Data is crucial for delivering accurate, compelling customer experiences and driving AI innovation. However, 90% of enterprise data exists in unstructured formats like PDFs, emails, social media posts, and audio files, making it largely inaccessible for business applications and AI models. Forrester predicts* that the volume of unstructured data managed by enterprises will double by 2024, highlighting the urgency of this challenge. While 80% of IT leaders acknowledge the transformative potential of generative AI in leveraging data more effectively, 59% still need a unified data strategy to harness this power.

SOURCE: Businesswire