The Advanced Dataset Creator

Most lists come together through search, yet some audiences depend on logic that search cannot express, such as counting, comparing, or computing across the data. The Advanced Dataset Creator handles those cases by building the list with SQL, the query language behind the dataset. You do not need to write the SQL yourself, because the agent proposes it from your description and waits for your confirmation, which is why this tool is now open to every user.

Companies where the engineering team is more than half of total headcount, in US SaaS

When search is not enough

Reach for the Dataset Creator when your audience depends on a calculation rather than a stored value. Counting how many of a role a company has, comparing two numbers, or computing a ratio are all things search cannot phrase but SQL can.

Counts and precise conditions

Aggregations and exact filters let you build lists that search cannot express, such as companies with more than a set number of a given role, or accounts that meet several precise conditions at once. This is where a question about how many, or exactly which, becomes a list.

Computed columns

You can add columns the dataset does not store directly, such as revenue per employee or the share of headcount in engineering. The Dataset Creator computes these from existing fields, so you can target on a ratio rather than a raw value.

Propose, then confirm

You describe the audience, the agent proposes the query, and you confirm it before it runs. That propose-then-confirm flow is why the tool is open to every user, not only people who know SQL. You read the query in plain terms and approve it, rather than writing it.

Save as a session

Save a dataset as a session so you can run it again as the underlying data changes. A definition you trust becomes a list you can refresh, instead of a one-time pull you have to rebuild.

Download the result

When the dataset is ready, download it for use in your other tools, or carry it forward into enrichment and execution inside Landbase.

Try it in Landbase

1. Think of an audience you could never quite build with filters, such as companies whose engineering team is more than half their headcount.
2. Describe it to the Advanced Dataset Creator and review the query it proposes.
3. Confirm the query and review the result.
4. Save it as a session so you can run it again next month.

Prefer the CLI? Describe the same dataset from your terminal:

landbase-cli search "companies where the engineering team is more than half of total headcount, US SaaS"