Is Data Cloud the same as Data 360?

Same platform, newer name. If a proposal treats them as two products, ask why.

Do we need it before Agentforce?

You need grounded, permissioned context. Sometimes that is Data Cloud. Sometimes it is clean Knowledge and a sharing model that already works. We check the org before quoting.

Does it replace the warehouse?

No. It resolves and serves the profile. The warehouse stays the analytical store, and we front it rather than move it.

Can it be the only memory layer?

For profile and similarity, often yes. For multi-hop relationships and decision history, put a graph next to it.

Salesforce Data Cloud

Data Cloud is the profile the agent grounds on.

Data Cloud resolves the customer across the systems that each think they own them, then serves that profile to journeys, analytics, and agents. It is context, not agency.

Get the written assessment Memory plane

Founded 2022 · Founder in the Salesforce MVP Hall of Fame · Dubai HQ · Kochi studio

In the org

Four pieces that matter to an agent.

Data Cloud is also sold as Data 360. Same platform, same four pieces we care about when an agent has to be right.

01
Data model objects
Canonical shapes for customer, account, case, product
02
Identity resolution
One profile · Match rules · Survivorship
03
Data spaces
Tenant isolation · Team scoping · Row-level governance
04
Hybrid retrieval
Structured fields · Vector search on unstructured text
05
Consumers
Agentforce · Marketing journeys · Analytics

Data Cloud for agents. Sources land in data model objects, identity resolution produces the profile, data spaces keep tenants and teams apart, and retrieval serves both structured fields and semantic search.

Mermaid source

flowchart TD
  A["Sources: CRM, ERP, web, documents"] --> B["Data model objects: canonical shapes"]
  B --> C["Identity resolution: one profile"]
  C --> D["Data spaces: tenant and team isolation"]
  D --> E["Retrieval: fields plus vector search"]
  E --> F["Agentforce, journeys, analytics"]

Why it matters

An agent without a profile guesses.

These are the four failures we find when an agent pilot is “almost right”.

Four records

One resolved profile

The agent answers about whichever record it found first: One customer, with match rules someone can review
Grounding ignores who may see what: Data spaces keep each desk inside its walls
Documents sit outside the profile, so nothing is cited: Fields and document text retrieved together, with citations
Limits are discovered in production: Ingestion and usage limits designed for up front

In an org (example)

A bank, and one customer with four records.

Step 1
Bring the sources
Core banking, the CRM, the servicing portal and the complaints inbox land as data model objects, not four bespoke syncs.
Step 2
Resolve the person
One profile, with match rules written down and reviewable.
Step 3
Keep the walls
Data spaces separate business units, so one desk's agent cannot browse another's.
Step 4
Serve the agent
Agentforce grounds on the profile. A code agent reaches the same context through the gateway.

Who this is for

Three desks with one customer, many records.

Sometimes clean Knowledge and a working sharing model are enough. We check first.

Example
Head of customer experience at a Riyadh telecom
Today: Billing, the app and the call centre each hold their own version of the subscriber.
After: The call-centre agent and the app both read one profile, with the match rules on file.
Example
Data lead at a Dubai real estate developer
Today: A buyer, a tenant and a lead are three records for one person.
After: One profile across sales, leasing and service, so marketing stops pitching units the person already owns.
Example
Agent programme owner at a Boston insurer
Today: A service agent pilot cites the wrong policy because it found the stale record first.
After: The agent grounds on the resolved profile and cites the live policy.

Examples of the work, not case studies. Roles and cities are typical of the desks we see.

How Mindcat helps

Context first. Then the agent.

We will tell you when clean Knowledge and a working sharing model are enough, and Data Cloud is not the first cheque to write.

Questions

What we actually say.

Is Data Cloud the same as Data 360?: Same platform, newer name. If a proposal treats them as two products, ask why.
Do we need it before Agentforce?: You need grounded, permissioned context. Sometimes that is Data Cloud. Sometimes it is clean Knowledge and a sharing model that already works. We check the org before quoting.
Does it replace the warehouse?: No. It resolves and serves the profile. The warehouse stays the analytical store, and we front it rather than move it.
Can it be the only memory layer?: For profile and similarity, often yes. For multi-hop relationships and decision history, put a graph next to it.

The brief · one email

Name the systems that disagree about a customer.

We will tell you whether Data Cloud is the fix or a bigger bill. The assessment is free: a short read of the org and a written note. If you go ahead, the first build week is a fixed AED/SAR 14,900 (about USD 4,050), credited in full when you sign the build within 60 days. The honest no: if the work is a software factory with a named delivery date, we are the wrong partner.