Salesforce Data 360 (Data Cloud) Implementation Services
Unify customer data once, then use it everywhere: in journeys, in dashboards, and as the context your AI agents reason over.
Atrium is a Salesforce consulting partner specializing in Data 360, the platform Salesforce previously called Data Cloud. Our implementation services cover ingestion and Zero Copy federation, harmonization and identity resolution, calculated insights and data graphs, segmentation and activation, governance and Data Spaces, and grounding for Agentforce agents.
Engagements start with a use case discovery workshop or a proof of value, then run through implementation and managed services.
What Is Data 360?
Salesforce Data 360 unifies customer data into governed profiles and objects that CRM, analytics, marketing, and AI agents all read from. Salesforce renamed it from Data Cloud, and Salesforce is a Leader in the Gartner Magic Quadrant for Customer Data Platforms for the third consecutive year.
Mechanically, data lands in Data Lake Objects, maps into Data Model Objects on the Customer 360 Data Model, and unifies into profiles through identity resolution. Sources arrive through the connector catalog, the Ingestion API, MuleSoft, or the web and mobile SDKs, or they stay where they are through Zero Copy federation.
Every organization we meet has already unified its customer data at least once: a warehouse, a CDP bought for one campaign, and reports that disagree. We design the harmonized model and identity resolution first, then build the graphs, insights, and segments your marketers, dashboards, and agents each call.
Get Your Data Working Everywhere
Two decisions set the shape of a Data 360 program: how identity resolves, and which sources are worth landing in the platform at all.
All six engagements below exist to get those two right early, because Data Spaces, consent checks and identity rules are cheaper to design than to retrofit.
What Ships Inside Data 360
The reference layer, in four segments. Open the one you are scoping.
Data streams bring batch, streaming, and incremental sources in through the connector catalog, the Ingestion API, the MuleSoft Anypoint Connector, or the web and mobile SDKs, landing in Data Lake Objects. Ingestion itself is free. Mapping source fields to the Customer 360 Data Model produces Data Model Objects, and identity resolution rulesets, built from match rules and reconciliation rules, resolve people, accounts, and households into unified profiles. Real-time identity resolution depends on data graphs.
Calculated insights run as scheduled SQL or through the no-code Visual Insight Builder, materialized so segments, activations, and data graphs reuse them. Data graphs are denormalized, materialized views that return full customer context in one sub-second call. Data transforms reshape data in batch or streaming form, and data actions fire platform events, webhooks, or Marketing Cloud Engagement data extensions from a streaming insight.
Standard, real-time, waterfall, dynamic, and nested segments are built on Segment Canvas or from a prompt, then activated to Marketing Cloud Engagement and Personalization, ad platforms, file storage, and API activation. Data Spaces are logical partitions separating brands, regions, or business units, with filtered objects, mappings, and permission sets, alongside field and record-level policies, masking, tagging, lineage, and consent checks.
Unstructured data support adds vector and hybrid search over documents, with Intelligent Context configuring the pipeline through prompts. Predictive AI and bring-your-own-model options cover scoring, Data 360 Clean Rooms cover collaboration, and data shares push to Snowflake, BigQuery, SageMaker, and Databricks. Data Cloud One connects multiple orgs to one home org, Tableau Semantics carries governed metrics into Tableau Next, and data kits and bundles make it portable.
Do You Have to Move Your Data into Data 360?
No. Data 360 supports Zero Copy federation alongside physical ingestion, with query and file federation and three access modes: physical ingest, Live Query, and Cached Acceleration. It runs on Apache Iceberg, and federated data is processed in memory and discarded after use.
Availability is source-specific, and this is where an architecture gets decided. Snowflake, Amazon Redshift, Google BigQuery, Databricks, IBM watsonx.data, and Apache Iceberg tables support Zero Copy. AWS Glue Data Catalog federation reached general availability in June 2026. Microsoft Fabric OneLake federation was in beta as of June 2026, with Fabric ingestion generally available in July 2026. Azure Synapse is batch only.
Real-time identity resolution, segmentation, and agent grounding all want data in the platform, so most architectures end up hybrid on purpose. Because Atrium also delivers Snowflake consulting, we design both sides of that connector rather than one.
Grounding is where agent projects are won. Structured grounding comes from data graphs, unstructured grounding from a search index configuration that parses, chunks, and embeds documents into retrievers, and semantic grounding from Tableau Semantics, so “churn rate” resolves to your official calculation rather than a model’s guess. Supported unstructured sources include Amazon S3, Google Drive, Microsoft SharePoint, Confluence Cloud, Zendesk, and a web crawler.
Why Atrium for Data 360
Data engineering and data science in one practice
Atrium builds the harmonized model, the identity resolution rules, the predictive models, and the analytics layer with one team, so nobody hands off a model nobody can score.
800+ Salesforce certifications across 275+ employees
Atrium holds seats on multiple Salesforce Partner Advisory Boards, including Data 360, Tableau, Sales Cloud, and Service Cloud, so the practice sees roadmap direction before it ships.
We work on both sides of the connector
Atrium delivers Snowflake as well as Salesforce, so the Zero Copy conversation is not one-sided. We define outcomes at the start and measure against them with conservative attribution, the method behind the $1B in measured customer impact figure.
Data 360 FAQ
Is Data 360 the same thing as Data Cloud?
Yes. Data 360 is the current name for the platform Salesforce previously called Data Cloud, renamed from Data Cloud. Salesforce still uses the older name in places, including some URLs and documentation, and the Winter ’26 release note announcing the change is titled “Data Cloud Is Now Data 360.”
Do we have to move our data into Data 360?
No. Data 360 supports Zero Copy federation alongside physical ingestion, with three access modes: physical ingest, Live Query, and Cached Acceleration. Snowflake, Amazon Redshift, Google BigQuery, Databricks, IBM watsonx.data, and Apache Iceberg tables are supported. Real-time identity resolution, segmentation, and agent grounding all work best with data in the platform, so most architectures end up deliberately hybrid.
What is identity resolution and why does it matter so much?
Identity resolution rulesets, built from match rules and reconciliation rules, resolve records into unified profiles for people, accounts, and households. It matters because everything downstream inherits it: segments, calculated insights, data graphs, activations, and any agent grounded on profile data. Getting the rules wrong is the most expensive mistake to correct late.
What does a Data 360 implementation actually deliver?
Use cases running in production, rather than a configured platform waiting for someone to adopt it. In practice that means profiles your teams trust, segments and activations working against them, and one model your dashboards and your agents both read from. Atrium sequences delivery so something is live and measurable early, which is also how you find out whether the identity rules are right before everything downstream depends on them.
How does Data 360 ground Agentforce agents?
Three ways. Structured grounding comes from data graphs, unstructured grounding from a search index configuration that parses, chunks, and embeds documents into retrievers, and semantic grounding from Tableau Semantics so business terms resolve to your official definitions. Supported unstructured sources include Amazon S3, Google Drive, Microsoft SharePoint, Confluence Cloud, Zendesk, and a web crawler.
Why does it matter that Atrium delivers Snowflake as well as Salesforce?
Because the decision about what to federate and what to land shapes how much you maintain and how fast it performs, and advice from one side of that connector tends to favour that side. Atrium builds on both, so the recommendation follows your latency and governance requirements rather than the platform we would rather sell. See our Snowflake consulting practice.
Data 360 Use Case Discovery Workshop
Not sure where to begin? This hands-on workshop is ideal if you’re just starting your Data Cloud journey, including those who want to reap the benefits of Salesforce’s free Data Cloud credits.
In 1-2 weeks, we’ll work with you to identify use cases that will have the fastest business impact. We’ll provide a custom agenda focused on use case ideation and identification, as well as develop personalized next steps. This workshop typically leads into Data Cloud MVP, which can help you quickly get up and running.
Latest Insights from Our Experts
Ready for Real Results?
Reach out to our team—dedicated to the metrics that matter most. We’re here to help you adapt and grow.