What’s New With Data Cloud? Sub-Second Processing, Data Formats, and More

Data Cloud has been in the limelight over the past few years, having graduated from a CDP for marketing use cases, subsequently changing names (remember Genie?), and now the data foundation that underpins every Salesforce ‘cloud’ product. 

Agentforce, Salesforce’s latest “hot” innovation, allows organizations to build and manage autonomous agents for tasks across various business departments. Salesforce underlines how key Data Cloud is to power these agents, feeding plenty of quality data to Atlas, the ‘brain’ behind Agentforce. After all, the quality of AI is only going to be as good as the data that underpins it.

With Data Cloud at the heart of Salesforce’s Customer 360 (application layer) and Agentforce, here are the new announcements that will give Data Cloud customers a competitive edge.

What’s New for Data Cloud? 

As Data Cloud continues to gain traction with customers, these three announcements may seem trivial but in fact are important for catering to the performance and scalability that organizations will need. 

• More Unstructured Data Formats Accessible: Native processing is now possible for video and audio formats.
• Additional Connectors: There are 50 new pre-built connectors launching.
• Sub-Second Real-Time Capabilities: Data Cloud will be able to work incredibly fast across all of the processes it handles. 

The image below indicates a) the three capabilities highlighted above, and b) how the Atlas engine works in tandem. 

More Unstructured Data Formats Accessible (Video and Audio)

The Vector Database earlier this year brought attention to unstructured data, followed by enhanced support for semantic data since then. With 90% of company data being unstructured (as opposed to structured in records, objects, and object relationships), companies are missing out on a wealth of information that can’t be surfaced in traditional reporting. 

With this announcement, native processing for video and audio (e.g. sales demo calls, customer service calls, voicemails) is possible, meaning that these can be analyzed to reveal insights that can be translated into usable data, eventually for segmentation and activation. 

New Data Cloud Connectors Available: Understanding the 200+ Collection

There are 50 new pre-built connectors launching. The new total will now be 200+ pre-built connectors, which take out much of the legwork when connecting disparate data sources. This ensures Agentforce agents are basing their reasoning and actions on a more holistic set of data.  

These connectors can pull data into Data Cloud and unify it with data from other sources. For example, pull in inventory data that resides in a third-party system. As a result, customers are given the right inventory information and can fulfill orders faster.

• “Horizontal”: Generic connectors, such as the web and mobile SDKs. 
• “Hyperscale”: More robust connectors for organizations working with high volumes of data, such as AWS or Kinesis. 
• Industry-specific connectors: Catering to industry use cases, for example, a medical device supplier who wants to reach into IoT (device usage) data.

Sub-Second Real-Time Capabilities

Data Cloud will be able to work incredibly fast across all of the processes it handles, from data ingestion to activation, all in sub-seconds. In my opinion, this is the most important of the three, reducing lags between these stages – you could say it can all happen faster than ‘a blink of the eye’. 

Data Graphs enable you to visualize the relationships between data model objects (DMOs), even going several layers deep. This helps you to investigate whether the correct data is available at the time of writing a prompt/instructing an Agentforce agent, which will be required to generate an optimal and accurate output. Real-time Data Graphs perform faster identity resolution, segmentation, and actions based on ingested data to ensure data is ready for Agentforce to work with. 

Zero Copy Integrations

The zero-copy partner network delivers connectors to other platforms in a way that data isn’t copied and stored in both systems; instead, it’s fetched when it’s needed and virtualized in Data Cloud. Data Kits are available, which are packages of metadata and components that speed up the installation process – as Salesforce say, “without having to write a single line of code”.

File Federation for Zero Copy Integrations

Pilot Oct ’24

With File Federation, you can securely access and analyze large data sets from your data lakes, ensuring that you’re staying compliant. Using zero copy integrations, you also boost your system’s performance as data is virtualized (i.e. not having to physically move data in order to act on it in Data Cloud). 

Data Cloud One

GA Oct ’24

Using zero-copy technology, you can implement a central Data Cloud org across multiple Salesforce orgs, accelerating your unified customer view without ever having to write a single line of code.

New in Data Security and Governance

Different enterprises have unique security needs, and with the volumes of data pumping through Data Cloud as hundreds of thousands of columns and tables, managing security and governance is inevitably challenging. Now, ensure that everyone is ‘reading off the same page’ when it comes to who should access what data. The Data Governance tab opens up these four enhancements, built natively into Data Cloud.

AI-Based Tagging & Classification

Beta Nov ’24

Use tags to classify data – for example, confidential data, security data, or financial data – which are applied to columns in data tables. This tagging is the basis for policy rules (covered next). 

There are two methods available to do tagging – Tagging Manager and Trigger Auto-tagging, which uses AI. The tags summary shows the tags themselves, the number of tagged objects, tagged fields, and related policies by data set. 

Policy-Based Governance

Beta Nov ’24

Define access control policies with policy rules. These are based on the tagging and classification of attributes (i.e. the information used to describe data model objects which are contained in fields in the Data Cloud app). These rules determine who can access which types of data, e.g. Lauren doesn’t have visibility into financial data. Data masking for sensitive data is also part of this. 

Private Connect

GA Nov ’24

Establish a secure, private connection between Data Cloud and your organization’s data warehouses. 

Customer Managed Keys

GA Now

Customer-managed encryption keys (keys that you own) will give you more control over when data is encrypted to ensure that it’s secure at rest (i.e. while stored). 

Data Cloud Roadmap

The momentum around Data Cloud is phenomenal. This extensive list is a good visual to demonstrate how much development resource is going into Data Cloud – including what’s become generally available in the past few months and what’s coming up on the roadmap in the next six months. 

We’ve picked out four that we think are the most notable:

• Data Cloud in Sandboxes: Build out and test Data Cloud configuration in an environment that’s not live. This is going to be a welcome update for the technical people out there (GA Nov ’24).
• DevOps Center for Data Cloud: DevOps Center is the place to manage releases and to give low-code and pro-code development team members a single set of configuration and code to work with bringing closer collaboration (GA Nov ’24).
• Privacy Center for Data Cloud: An extension of the security and governance features now in beta, Privacy Center handles sensitive data management (anonymize, de-identify, or blank), data requests, and consent management (Dec ’24, Pilot or GA not confirmed). 
• Communication Capping: Similar to features in Marketing Cloud, the assumption is that this will place guardrails on how many messages can be sent to customers within a given period of time (GA Dec ’24). 

Data Cloud Use Case Library

There’s a new location for exploring industry-specific use cases (although, most are not exclusive to one industry). Filter by industry or products you have available in your Salesforce stack (for data ingestion and activation). Each entry lists out the data sources used, suggestions calculated insights and predictive models, and the expected business benefits.   

Visit the library here. 

Datablazer Community

A new community will bring together IT, developers, and data practitioners that are Data Cloud enthusiasts. This is hosted on the Trailblazer community, where members will discuss Data Cloud strategies and solve each other’s queries. Scan the QR code below to join.

Summary

As Data Cloud continues to gain traction with customers, and is now playing a fundamental role for Agentforce, these three announcements are important for catering to the performance and scalability that organizations will need. 

One interesting aspect to mention is that Salesforce are finding that there’s a fruitful feedback loop occurring within their product development. Customers using Data Cloud are finding innovative ways to utilize this flexible platform, and in turn, this has prompted Salesforce to build new capabilities or forge partnerships with other vendors to deliver pre-built connectors. 

One example mentioned was when Salesforce started looking into sub-second real-time capabilities. Their first instinct was to base this requirement around the web and mobile SDK to bring in clickstream data, which then expanded to the popular IoT data use case. This is a two-way learning exercise that will continue to give Data Cloud customers a competitive edge.