Skip to Content
Skip to Footer

Salesforce Announces General Availability of Data Cloud Vector Database, Unleashing the Power of Unstructured Data and Generative AI to Transform Customer Experiences

Businesses can now unlock insights from previously inaccessible unstructured data and take action across all channels and touchpoints, including sales, service, marketing, commerce, analytics, and other applications

Salesforce today announced the general availability of Data Cloud Vector Database to help businesses unify and unlock the power of the 90% of customer data that is trapped in PDFs, emails, transcripts, and other unstructured formats. 

Businesses can now integrate unstructured data sources — such as call transcripts, online customer reviews, and support tickets — directly into customer profiles to gain a deeper understanding of customer needs and preferences, without having to rely on expensive and labor-intensive solutions. These enriched profiles enable teams to search through vast volumes of this data and to surface insights and content that can be used to enhance sales, service, marketing, and commerce experiences. For instance, Einstein AI can use the data to proactively schedule service calls, help sales teams close pipelines faster, and drive more strategic marketing campaigns.

Data Cloud uniquely harnesses the power of Salesforce metadata in the Einstein 1 Platform by linking unstructured and structured data. With the power of this combined data, analysts can now explore and visualize in Tableau, developers can create Salesforce Flow automations, and business users can ground their generative AI prompts. It also reduces the need to fine-tune Large Language Models (LLMs) while improving the accuracy of results provided by Einstein Copilot, Salesforce’s conversational AI assistant for enterprises.

 Salesforce’s EVP & GM, Unified Data Services & Einstein, Rahul Auradkar explains Data Cloud Vector Database.

Diving deeper

Integrated into Einstein 1 Platform, Data Cloud Vector Database ingests, stores, unifies, indexes, and allows semantic queries of unstructured data to take advantage of knowledge across all applications. 

Data Cloud can now ingest disparate unstructured content from customer interactions across multiple touchpoints, including websites, social media platforms, and commerce channels, at scale. It then uses the power of generative AI to create embeddings on unstructured data that are indexed in the vector database. Additionally, businesses can choose to include relevant structured attributes and semantically query the vector database. This enhanced data retrieval capability powers prompts and copilots to deliver more relevant, accurate, and up-to-date responses.

One of the strongest examples of Data Cloud Vector Database in action is how it enriches AI prompts with customer data to be used anywhere on the Einstein 1 Platform. In this video, a sales manager asks Einstein Copilot to create a meeting brief for an upcoming sales meeting.

Build and Close Pipeline Faster: Using Data Cloud Vector Database, customers can uncover new sales and service opportunities, such as:

  • Improving prospecting: Sales teams want to choose the best opportunities to pursue, make personalized sales plans, and proactively identify customers at risk of churn. However, traditional leads and opportunity scoring based on historical data available only in Salesforce lead to an incomplete picture of the prospect. With Data Cloud Vector Database, scoring factors include customer and product fit, past purchases, account scores, support interactions, usage patterns, and online interactions. Lead quality improves and sales teams can focus on the most promising opportunities.
  • Responding quicker to sales RFPs: Sales teams want to use Einstein Copilot to help craft responses to a prospect’s request for proposal (RFP) and win potential deals on time. However, the current recommendations do not consider the vendor’s capabilities that exist in vendor documentation. With Data Cloud Vector Database, Einstein Copilot taps into previous RFPs and other vendor data in knowledge articles and white papers to provide accurate responses that showcase the vendor’s strengths. 
  • Personalizing outreach: Sales teams want to create personalized sales outreach. Yet, customer interactions and behaviors across the company aren’t fully considered when customer outreach emails are created. Einstein Copilot taps into Data Cloud Vector Database and uses knowledge articles, PDFs, account history, and other unstructured data to write emails that are personalized for each customer, increasing the chances of closing sales.

Enhanced Service and Support:

  • Personalizing customer engagement: Service teams strive to know their customer preferences, predict their needs, and offer tailored services. However, current customer profiles typically only include basic details like name, account details, and past tickets. With Data Cloud Vector Database, customer profiles will be enriched with behavioral data, preferences, and purchase histories, enabling services teams to give customers hyper-personalized care. 
  • Efficiently managing knowledge: Service agents and bots need to provide faster, more personalized responses to customer requests. However, much of an agent or bot’s time is spent searching for the right knowledge article to find the right solution. Data Cloud Vector Database, on the other hand, understands the context, connections between articles and tickets, and customer history, enabling agents and bots to find the most relevant troubleshooting tips quickly and precisely. 
  • Improving cross-sell and upsell recommendations: Most service teams want to strengthen customer relationships and boost revenue by offering tailored product suggestions to existing customers. Yet, current product suggestions often overlook customer needs and preferences. Data Cloud Vector Database enables AI to suggest intelligent cross-sell and upsell product recommendations to agents, based on customer preferences, website interactions, social media engagements, and order histories. 
  • Proactively resolving issues: Service teams try to identify potential issues before they escalate into significant problems. But they don’t always spot patterns that point to emerging issues, equipment failures, or other oncoming disruptions. Data Cloud Vector Database can proactively manage equipment and assets, taking into account details like age, usage, and repair history to calculate an Asset Health Score and act to automatically plan service appointments, find and fix problems, and recommend upgrades for aging assets.

Customer perspective

“The integration of Data Cloud Vector Database with Einstein for Service and Prompt Builder yielded high-accuracy automated customer responses. As a bank which champions local commitment combined with strong central support, we are always looking to improve our response rate and precision in Spar Nord’s Contact Center.” – Kim Kirkeby, SVP Customer Service, Spar Nord

Salesforce perspective

“For the first time, business users can bring 90% of untapped data to their workflows, which has so far remained unstructured and unused in their enterprise. With the new innovation in Data Cloud, businesses can now grow cost effectively by delivering transformative and integrated customer experiences across service, sales, marketing, AI, automation, and analytics.” – Rahul Auradkar, EVP & GM, Unified Data Services & Einstein, Salesforce

With the new innovation in Data Cloud, businesses can now grow cost effectively by delivering transformative and integrated customer experiences across service, sales, marketing, AI, automation, and analytics.

Rahul Auradkar, EVP & GM, Unified Data Services & Einstein, Salesforce

Related news: In addition to the general availability of Data Cloud Vector Database, Salesforce today announced new capabilities to get even more value from AI and Data Cloud, including:  

  • Hybrid search, now in pilot, provides even greater search accuracy by combining semantic search with traditional keyword search. By understanding semantic similarities and context while also recognizing company-specific words and concepts, hybrid search enables customers to improve information discovery, and deliver more relevant and tailored search results from a large database of structured and unstructured data.  
  • Data Cloud in Sandbox Environments, available in Open Beta later this month, enables IT teams to configure and test Data Cloud innovations within existing Salesforce Sandboxes before deployment. All of Data Cloud’s metadata, including data streams, calculated insights, zero copy integrations, and search indexes, are available to be customized against existing Salesforce configurations and workflows. Additionally, Salesforce DevOps Center will support Data Cloud, starting as beta in July, integrating it into the existing application lifecycle management process.
  • Data Cloud on Hyperforce in the United Kingdom, available July 31, gives companies the ability to grow globally and serve customers locally. Organizations will be able to securely store data in-country, better meet global compliance standards, and experience the same real-time customer journeys and engagement at scale with local data storage.

Learn more:

Any unreleased services or features referenced here are not currently available and may not be delivered on time or at all. Customers should make their purchase decisions based upon features that are currently available.

Get the latest Salesforce News

Exit mobile version