Build Dataflow Dashboards in Python¶

panel-flowdash is a framework for building visual, dataflow-driven dashboards with Panel. Define components with typed ports, wire them together, and persist layouts to SQLite.

Why FlowDash?¶

• Component-first architecture: define reusable building blocks with @register or as Viewer subclasses.
• Typed dataflow: ports carry type information for compile-time and runtime validation.
• Visual wiring: connect components in a node editor, with cycle detection and single-source enforcement.
• Persistence: dashboard layouts, edges, and grid positions are stored in SQLite.
• One command to serve: flowdash serve my_project/ scans, registers, and launches.

Quickstart¶

from panel_flowdash import register

@register(component=True, provides=[{"key": "ticker", "type": "str"}])
def app(config):
    import panel_material_ui as pmui
    return pmui.Select(name="Ticker", options=["AAPL", "GOOG", "MSFT"])

flowdash serve my_project/ --port 5006

How-to guides¶

• Register Components - the @register decorator and Viewer classes
• Define Ports - typed inputs and outputs for dataflow wiring
• Wire the Dataflow - connecting components with validation
• Persist Dashboards - SQLite storage for layouts and edges
• Serve a Project - the flowdash serve CLI

Examples¶

• Examples gallery

Reference¶

• API reference
• Release notes