Skip to main content

Evaluating BimlFlex

This guide gives you a structured path for evaluating BimlFlex. It is organized around milestones, not a calendar: what to do in your first five minutes, what to work through in your first five days, and where to go deeper in your first five weeks. Every step in this guide uses surfaces that are built into the product, so you can follow along inside the app as you read.

Before you start, install BimlFlex and its metadata database using the installation guide.

Your first five minutes​

The goal of your first five minutes is to load a working sample and see a complete warehouse model, before you configure anything of your own.

When you open BimlFlex with an empty workspace, the dashboard asks "Where would you like to start?" and offers two starting points, plus a path for your own data:

  • Build a dimensional warehouse (recommended): load a working sample, then follow the guided walkthrough to a published star schema.
  • Explore Data Vault automation: load a sample and see a full Data Vault warehouse end to end.
  • Have your own source data? The Import Metadata link takes you straight to importing metadata from your own sources. The first five days section covers that path.

Both starting points lead to the Sample Gallery, which loads a working sample into your selected version. The gallery groups samples into three tiers:

  • Starting points: sample connections, batches, and projects for a specific target platform and orchestration style, ready for you to import source metadata and model it.
  • Complete solutions: the finished metadata for a whole warehouse project, ready to explore, trace, and build.
  • Guided journeys: one case study captured at successive stages of delivery, so you can join the story at the step you care about and compare your work against the finished layers.

A Recommended row at the top of the gallery highlights the best entry points. Every card has two actions: Preview, which shows the sample's description, entity counts, an estimated time, and its suggested first stops, and Load sample, which loads it into your workspace.

Get oriented on Sample Home​

After a sample loads, BimlFlex opens Sample Home. It tells you exactly what was loaded and where, gives you a short orientation for the sample, and lists "What to look at first" links into the model. From here you choose your modeling journey:

  • Dimensional flow: model a star schema from the loaded metadata with the Dimensional Accelerator and its guided walkthrough.
  • Data Vault flow: accelerate the loaded metadata into a Data Vault model with hubs, links, and satellites.

Sample Home also carries the first-build checklist (covered below), a build readiness strip with an Open Build entry point, and actions to capture a checkpoint or reload the sample, so you can experiment freely and get a fresh copy at any time.

The recommended evaluation journey is the guided Dimensional Accelerator. It walks you from source objects to a published star-schema model in four steps: pick a target project, select your sources, review the derived dimensional model, and publish. The walkthrough is built into the page, so your first pass through it is guided end to end, and the derivation gives you dimensions, facts, surrogate keys, and change-type defaults in one pass.

The Data Vault journey​

If Data Vault is your methodology, take the Explore Data Vault automation starting point instead. Its guided journey is built around the Willibald case study, a fictitious seed-and-plant retailer whose webshop orders and point-of-sale roadshow transactions come together in one integrated model. The journey is staged: source imported, source modeled, business model captured, Data Vault accelerated, and data marts derived on top of the vault. Load the stage you want to study, or load the final stage and trace the finished solution end to end.

Your first five days​

The goal of your first five days is to go from touring a sample to working with your own metadata and running your first build.

The spine for this stage is the first-build checklist, shown on the dashboard and on Sample Home. It tracks five steps against the live state of your workspace, and it expands the first step you have not finished yet:

  1. Connect your workspace. Pick your metadata database, customer, and version. This is the workspace everything else lands in.
  2. Get metadata in. Load a sample warehouse, or import your own source metadata. To import your own, open Import Metadata from the Design section of the navigation: select the source connection, use Preview to retrieve the object list, select the tables you want, and choose Import. If the source connection does not exist yet, you can create it without leaving the page.
  3. Model your warehouse. Follow the Dimensional Accelerator walkthrough (recommended), or accelerate a Data Vault.
  4. Check build readiness. Confirm the environment is ready to run a build before you press the button. The readiness strip checks your environment (licensing, the Biml compiler, bundle pairing, the workspace store, and workspace settings) and your metadata (connections and their source-to-target pairing, a project, imported metadata, and a modeled warehouse). Every check that does not pass links to the place where you fix it.
  5. Run your first build. Open the Build page and run the build. The same readiness strip sits above the build, so you always know the state of the environment you are about to build in.

Two product behaviors are worth knowing while you work through these steps:

  • Checkpoints. Capture a checkpoint before you try something big, and roll back from the dashboard's Snapshots list if you want your earlier state back. BimlFlex can also capture a checkpoint automatically before risky operations; see the AutoCheckpointBeforeRiskyOperations setting.
  • Validation guidance. As you edit metadata, validation findings do more than flag a problem: they carry guidance for resolving it, and where a setting or concept needs background, they link to the relevant documentation page.

Your first five weeks​

By this stage you have a modeled warehouse and a first build. The next depth comes from the surfaces you will rely on in real delivery work.

Understand the settings that shape your output​

BimlFlex's generated output is driven by settings, and the product is built to make them understandable. In the Settings editor, every setting shows its description and shipped default, an Explain this setting affordance breaks down what the current value does, and settings that do not apply to your platform and configuration are hidden by default (one click shows all, and inapplicable settings are badged when shown). Most settings also have a reference page in this documentation; three worth reading during an evaluation:

Trace lineage and assess impact​

Two lineage surfaces answer "where does this come from, and what does changing it touch": Data Lineage in the Design section, and Model Lineage in Build & Verify, which traces objects and columns across the whole generated warehouse. Lineage is column-level wherever your authored mappings, key compositions, expressions, or conservative column-name matching establish the link, with object-level flow everywhere else.

Capture the business model​

The Business Modeling section lets you describe your organization's concepts (models, entities, and a taxonomy) alongside the physical metadata, so the warehouse you accelerate reflects the business's language, not just the source schemas.

Generate documentation​

The Documentation page in Build & Verify (the same surface the toolbar's Documentation menu opens through its Generated documentation item) generates browsable documentation of your workspace from your metadata, including model diagrams and lineage. It is a fast way to share what you built with people who were not in the tool.

From evaluation to production​

When the evaluation turns into a delivery plan, the delivering-solutions guides cover the architecture layer by layer:

Where to get help​