Build Data Pipelines Visually, Compile to Real SQL

Real data workflows are not flat sequences. They involve shared transformations that multiple downstream pipelines depend on. Plotono supports pipeline composition natively: any pipeline can reference another pipeline as a node in its own graph.

You can reference a pipeline by its unique ID, by its human-readable name, or by embedding it inline. When you reference a pipeline, its parameters become available in the parent. Parameter renaming lets you map a child pipeline's @minimum_age parameter to the parent's @min_age, keeping interfaces clean without modifying the original pipeline.

Composition works at multiple levels. A pipeline can reference a pipeline that itself references another pipeline, and parameter renaming propagates recursively through every layer. The compiler detects circular dependencies at compile time and reports a clear error rather than entering an infinite loop.

Every visual pipeline compiles to standard SQL. Plotono supports both DuckDB and BigQuery as compilation targets, so the same pipeline definition can run against a local analytics database or a cloud data warehouse. The compiler follows a four-stage process: lexing (tokenization), parsing (AST generation), optimization, and code generation.

Plotono also provides a pipe syntax that sits between raw SQL and the visual builder. Pipe syntax lets you write queries as a chain of operations separated by the |> operator. The compiler transforms pipe syntax into the same internal AST that the visual builder produces, so optimizations apply equally regardless of which interface you use.

Twelve Query Optimizers

Before code generation, the AST passes through up to twelve optimization stages. Each optimizer targets a specific class of inefficiency:

The pipeline editor supports three modes that you can switch between at any time. Visual mode gives you the drag-and-drop canvas. Pipe syntax mode shows the pipeline as a chain of operations using the |> operator. Raw SQL mode shows the compiled SQL output, which you can also edit directly.

All three modes produce the same underlying representation. A pipeline built visually generates the same SQL as one written in pipe syntax or raw SQL. This means you can start with the visual builder to prototype quickly, switch to pipe syntax to fine-tune the logic, and then review the raw SQL to verify what will actually execute against your database.

Plotono's workspace system is designed for teams of any size. Each organization gets a tenant with strict data isolation. Within a tenant, workspaces form a hierarchical tree up to ten levels deep, letting you mirror your organizational structure as precisely as needed.

Role-based access control assigns every workspace member one of three roles: admin, editor, or guest. Admins can manage workspace settings and members. Editors can create and modify pipelines, visualizations, and dashboards. Guests can view shared resources but cannot modify them. Permissions can propagate to child workspaces automatically, or be scoped to a single workspace.

For access patterns that do not fit neatly into the workspace hierarchy, tag-based access control lets you create cross-cutting permission groups. Apply a tag to a pipeline, connector, or visualization, and grant access to that tag independently of workspace membership.

Once your pipeline is producing clean, structured output, you can map its columns directly to chart axes. The column mapping editor lets you assign columns to X and Y axes, color dimensions, size encodings, and labels. Choose from over twenty chart types and drop the resulting visualization onto a dashboard.

Because dashboards are connected to pipelines rather than static data exports, your charts update whenever the underlying data changes. Global dashboard filters bind to pipeline parameters, so a single date range picker or category dropdown can filter every chart on the page simultaneously.

This tight integration between pipeline building and visualization eliminates the disconnect that plagues traditional BI setups where the transformation layer and the visualization layer are separate products that communicate through database tables.