Flow

What is Flow

Flow is the central solution marketplace within the ProMediaFlow platform. It provides a catalog of packaged business workflows—referred to as "solutions"—that combine operational styles and automated processes to assist with your daily operations.

Workflows in Flow operate along an operational autonomy ladder, transitioning from human-supervised tasks to fully autonomous execution. As solutions run, the platform tracks operational performance metrics, allowing precision, quality, and completion speed to automatically improve over sequential executions.

Getting started

To open Flow and start working with workflow solutions:

1. Log in to your ProMediaFlow dashboard.
2. Select Flow from the main dashboard menu to open the marketplace catalog.
3. Browse standard platform templates or manage your customized workflows.

Note: Executing workflow tasks and generating AI assets consumes platform credits.

Key capabilities

• Solution Marketplace & Execution: Browse, launch, and execute autonomous business process workflows categorized by Type, Subcategory, and Style to automate daily operational tasks.
• System vs. Custom Filtering: Filter the solution catalog by origin to separate platform-provided standard templates (System) from your own copied or customized workflow instances (Custom).
• Maturity Badges: Evaluate workflow reliability at a glance using dynamic badges calculated from run history, quality scores, user acceptance ratings, and autonomy levels:
  • Verified: Workflows proven reliable through high run counts, strong quality scores, and high user acceptance.
  • Experimental: New, modified, or unverified workflows restricted to a supervised sandbox environment.
• Autonomy Scale: See where each workflow sits on the operational ladder, tracking its progression from manual, human-supervised execution to full autonomy.
• Process Map & Resources: Review the complete process architecture and sequential execution steps alongside the required AI capabilities and human roles needed for the workflow.
• Learning Curve: Track live performance trends across runs, including quality scores and time-to-task completion duration as a solution learns and improves over time.
• Version Selector: Inspect every historical model version of a solution to review how evaluation scores and performance metrics have evolved across iterations.

How to use it

1. Browse and filter solutions

Open the catalog and narrow down available workflows using the filter controls:

• Filter by metadata attributes: Type, Subcategory, and Style.
• Select System to display platform-provided standard workflow templates.
• Select Custom to view workflows created or modified by you and your team.

2. Inspect workflow details

Select any solution card in the catalog to view its overview:

• View the Process Map for a visual layout of sequential execution steps.
• Check Resources to review required AI capabilities and human roles.
• Examine the Autonomy Scale to see if the workflow requires manual supervision or runs autonomously.
• Use the Version Selector to compare current and past model versions and evaluation scores.
• Review the Learning Curve to analyze quality ratings and task duration over previous runs.

3. Fork a solution

To customize a standard platform solution:

1. Select a System solution card in the catalog.
2. Select the Fork option.
3. Flow creates an independent, editable Custom workflow instance detached from the original template.

4. Execute a workflow

When you run a solution, its execution environment depends on its maturity badge:

• Experimental Workflows: Execute inside a supervised sandbox environment to safely monitor task execution and manage risk.
• Verified Workflows: Run with higher levels of unsupervised autonomy based on demonstrated performance and user acceptance.

Tips & notes

• Continuous Learning Loop: Each execution records four performance metrics: run count, quality score, user acceptance rating, and time-to-task duration. This feedback loop automatically enhances future precision and execution speed.
• Sandbox Reset on Modification: Editing an existing workflow automatically resets its badge to Experimental. Modified workflows must undergo testing in the supervised sandbox environment before re-qualifying for Verified status.
• Credit Consumption: Running workflow tasks and generating AI assets consumes platform credits.