For the complete documentation index, see llms.txt. This page is also available as Markdown.

What Are Datasets in Explore?

Quick Answer: A dataset in Explore is a named table of your business data β€” sales, inventory, forecasts, costs, or purchase orders β€” that's kept up to date automatically from your connected integrations. Pick one (or several) on the Explore landing page to tell Mo exactly what to build your report from.

What Is a Dataset?

A dataset is a single, ready-to-use table of your business data that shows up under Available datasets on the Explore landing page. Each one represents one type of information β€” daily sales, current inventory levels, a forecast, purchase orders, and so on.

You don't create datasets by hand. They're kept current automatically, in the background, so that whenever you build a report in Explore, Mo is working from your latest numbers.

Where Do Datasets Come From?

Every dataset traces back to data you've already connected to Moselle β€” your sales channels, inventory systems, financial tools, and any files you've uploaded (see How Your Data Works in Moselle for how that connected data flows in and how sync timing works).

Behind the scenes, Moselle organizes that connected data into two kinds of datasets:

  • Standard datasets β€” core, always-on data like sales, inventory, and forecasts. These refresh on a regular schedule and are available to every Moselle company by default.

  • Custom datasets β€” additional data sets your company has asked Moselle to make available, tailored to your business (for example, a specific financial rollup or an operational view unique to your workflow).

Why Don't I See a Dataset I Expect?

A few common reasons a dataset might be missing:

What You're Seeing
Likely Cause

No datasets appear at all

Your company hasn't enabled any datasets yet β€” contact support

One dataset is missing, others show up fine

That dataset may not be enabled for your company, or is still syncing for the first time

A dataset appeared but looks out of date

It refreshes on its own schedule β€” check back later, or ask Mo when it last updated

Enabling and managing which datasets are available for your company is handled by a Moselle admin under Settings. If you think a dataset should be turned on, ask your admin or reach out to support.

How Datasets Shape What Mo Builds

Different datasets are built for different kinds of reports. Picking the right one (or combining a few) steers Mo toward the answer you're actually looking for:

Dataset
Good For Building

Sales

Revenue trends, channel breakdowns, sales velocity

Inventory

Stock coverage, weeks of supply, at-risk items

Item Projections

Forecast vs. actuals, cross-forecast comparisons, prior-year trends

Purchase Orders

PO allocation, receiving status, supplier lead time

Costs

COGS breakdowns, margin analysis

You're not limited to one dataset per report β€” tag more than one in your prompt (for example, Sales and Item Projections together) when you want Mo to combine them, like a forecast-accuracy report that lines up actuals against projections.

Example: Build a Forecast Comparison Dashboard

Say your team builds more than one forecast for the same period β€” a top-down forecast, a bottoms-up forecast, and a custom forecast someone put together by hand β€” and you want one dashboard that lines all of them up against each other, against actuals, and against prior year.

1

Tag the Item Projections Dataset

On the Explore landing page, click Item Projections under Available datasets to tag it. This dataset holds your forecasts and actuals, so once it's loaded, Mo can compare across any of them. If you don't see it listed, contact support@moselle.io to have it enabled for your company.

2

Describe the Comparison You Want

With the dataset tagged, ask Mo directly β€” for example: "Build me a dashboard comparing my top-down forecast, bottoms-up forecast, and custom forecast against actuals and prior year."

3

Review and Save

Mo builds the comparison as charts and tables on your canvas. Refine it conversationally, then save any card you want to keep to a report β€” see Save & Share Reports from Explore.

Picking a Dataset When You Build

1

Open Explore and Find Available Datasets

On the Explore landing page, look under Available datasets. Datasets are grouped into folders β€” click the arrow next to a folder name to expand it and see the tables inside.

2

Preview Before You Commit

Hover over a dataset to see its metadata β€” the fields it includes, like identifiers, country, or brand. Click Preview to open a flyout with a sample of real rows, so you know what you're working with before you build.

3

Tag It in Your Prompt

Click a dataset to add it as a chip directly in your prompt box, then describe what you want β€” for example, tag Inventory and type "show me weeks of supply by channel." Mo uses exactly the data you pointed it to.

Frequently Asked Questions

Do I need to pick a dataset before building a report?

Answer: No. Tagging a dataset is optional β€” Mo can find the right data on its own from your prompt. Tag one only when you want to be specific about the source.

What's the difference between a "dataset" and a "report"?

Answer: A dataset is raw, live data you build from. A report is the finished chart or table you build with Mo and save β€” see Save & Share Reports from Explore for what happens after you build.

Can I combine more than one dataset in a single report?

Answer: Yes β€” tag as many datasets as your report needs, and Mo will pull from all of them.

How current is the data in a dataset?

Answer: Standard datasets refresh on a regular schedule, and custom datasets follow whatever cadence was set up for your company. Every report also re-runs against the latest data each time you open it, so you're never looking at a stale snapshot.

🧭ExploreBuild Charts & Tables in ExploreHow Your Data Works in Moselle

Last updated