Tools

Data 360 tools.

Interactive visuals and simulations that make the core ideas behind Salesforce Data 360 (formerly Data Cloud) easy to see, and easy to explain to someone else.

Most of what makes Data 360 valuable happens where nobody can see it. Records from different systems get mapped into one shape, then matched into one person. When those steps go right, every report, segment, and AI agent downstream sees the whole customer. When they go wrong, everything still looks fine. It's just quietly wrong.

The first two tools show both steps with fictional data you can poke at, and they work best in order: first harmonization, then identity resolution. The others cover what sits around them: why an activated audience comes out smaller than its segment, how data from an external data lake gets into Data 360 at all, and how documents become answers an agent can use.

  1. Step 1 Data harmonization, visualized Follow one customer across a CRM, a web store, a loyalty program, and in-store sales as they map onto one data model. Switch off a step and watch what breaks.
    • Why contact points and loyalty IDs live in their own objects
    • How an unmapped field turns real orders anonymous
    • Why an AI agent gives a confident, wrong answer on bad data
    Open the visualizer →
  2. Step 2 Identity resolution simulator Build match rules and watch records regroup into unified profiles, with every match explained. Use the demo data or load up to 50 of your own records.
    • How match rules combine, and why fuzzy names need a second signal
    • Why strict rules suit billing and relaxed rules suit marketing
    • How to spot over-merged and split profiles before they reach production
    Open the simulator →
Also: getting the audience out Why activations shrink A segment of 40 activates to fewer, or sometimes more. Change the membership, channel, source priority, and consent filters, and see which email each person actually gets. Open the visualizer → Also: connecting your data lake External data lakes and zero copy Ingestion, query federation, cached acceleration, and file federation with a lake such as Snowflake or Databricks: what each does, when to use it, and how to find the most cost-effective option. Open the guide → Also: grounding agents on documents Vector and hybrid search How a vector database works in Data 360: chunk real help documents, see them as embeddings, then watch vector, keyword, and hybrid search decide what an agent gets to read. Open the visualizer →

All sample data is fictional. Anything you load stays in your browser and is never uploaded or stored. These are independent teaching tools, not affiliated with Salesforce, and the logic is simplified for learning.

Want the reasoning behind the rules? Start with planning for identity resolution, the keychain analogy, and why Data 360 doesn't build a golden record.