Skip to content

Deployed solution walkthrough

  • Audience: Retail, analytics, and technology stakeholders
  • Duration: 3-5 minute orientation; 15-25 minutes for the full walkthrough
  • Data: Synthetic

Use this guide for a browser-first tour of a deployed Microsoft Fabric Retail Demo workspace. It complements the presenter demo, which contains the detailed talk track and support boundaries.

Representative screenshots

The screenshots show a representative deployment. Fabric navigation, generated values, item counts, and data timestamps can vary by deployment and run date. Validate freshness before presenting any value as current.

Before you start

  1. Complete Getting started.
  2. Open the deployed Fabric workspace.
  3. Confirm that retail_lakehouse, retail_eventhouse, the KQL queryset, retail_model, the pipelines and notebooks, and any optional ontology or agent items you plan to show are available.
  4. Run the historical setup notebooks. Run stream-events.ipynb only when the walkthrough needs recent Eventhouse data.

1. Orient the audience in the workspace

Open the workspace task flow and select the Retail Demo flow.

Task flow is a full-demo experience

The task flow is published only after the full-demo ontology step is complete. If the workspace uses core or standard, skip this step and open the workspace item list. You can still follow the historical, live, and Reporting paths described below.

Retail Demo workspace task flow

The task flow groups ingestion, transformation, storage, semantic, ontology, ML, and agent assets into one workspace-level map. Required ML points into the Semantic Model; post-Reporting optional and experimental ML points out from it so those extensions are not presented as Reporting prerequisites.

Use the flow to explain three paths:

  • Historical: load and transform deterministic retail history into the Lakehouse.
  • Live: send typed retail events to Eventhouse and optionally project them into Lakehouse tables.
  • Serving: expose curated data through the semantic model, Power BI, ontology, and data-agent surfaces.

The task flow is a navigation aid. It does not prove that a notebook, pipeline, or stream completed successfully.

Continue the walkthrough

Choose the page that matches the audience:

Walkthrough Use it for
Data platform Pipelines, Spark notebooks, Lakehouse history, and Eventhouse/KQL
Analytics and AI Ontology, grounded data-agent answers, and the Power BI report

For the complete demo, present the data platform first and finish with analytics and AI. For a business audience, use this overview and go directly to Analytics and AI.

Shared support boundaries

  • Screenshots show a representative deployment; item counts, values, and timestamps vary.
  • A task-flow node or pipeline canvas is navigation context, not execution evidence.
  • Validate the selected data period before discussing any value.
  • Skip optional ontology, agent, ML, or live-streaming surfaces that have not passed their source, permission, and capability checks.

What to check during the walkthrough

What you open What a ready workspace should show
Open the Retail Demo task flow Historical, live, storage, semantic, ontology, ML, and agent tasks are visible.
Open the Data platform guide Pipeline, notebook, Lakehouse, and Eventhouse steps are available without analytics detail.
Open the Analytics and AI guide Ontology, agent, and report steps are available without implementation detail.

For a timed presentation, continue with the presenter demo or a focused presenter journey.