ProductFlow is an open-source, self-hosted workbench designed for individual merchants and small teams to streamline product material creation and visual workflows. It integrates AI for generating copy, posters, and managing image sessions, offering a centralized platform for visual product assets.

How It Works

The system employs a backend API (FastAPI) orchestrated with PostgreSQL for data persistence and Redis for task queuing via Dramatiq workers. It features a visual, node-based canvas for organizing product data, reference images, AI-generated copy, and image generation nodes. Users can configure various AI providers, including mock, OpenAI, and OpenAI Responses, to generate content and manage image sessions with history tracking.

Quick Start & Requirements

Self-hosting is primarily achieved via Docker Compose, requiring only Docker/Docker Compose on the host. Key dependencies include PostgreSQL and Redis, managed within the Compose setup. Initial setup involves copying and configuring environment variables (.env.example) for administrative access keys and database credentials. Local development requires Python 3.12+, Node.js 20+, uv, pnpm, and Docker. Official documentation is available for user guides and architecture details.

Highlighted Details

• Visual Workflow Canvas: An interactive, zoomable, and pannable canvas for designing product workflows with drag-and-drop nodes.
• AI Content Generation: Supports AI-driven copy generation and image creation for posters and visual assets, with configurable models and parameters.
• Continuous Image Sessions: Enables iterative image generation with history, base image selection, and multi-candidate comparison.
• Configurable AI Providers: Integrates with mock, OpenAI (images/text), and OpenAI Responses image generation tools.

Maintenance & Community

No specific details regarding active maintainers, community channels (e.g., Discord, Slack), or sponsorship were found in the provided README.

Licensing & Compatibility

The project is released under the MIT License, offering broad compatibility for commercial use and integration into closed-source projects.

Limitations & Caveats

The current deployment is a single-administrator instance, lacking multi-user/multi-tenant support, team permissions, or payment systems. It does not offer production orchestration packages like Kubernetes or Helm. Deletion of products and image sessions is disabled by default to preserve audit trails, requiring explicit administrator enablement. Video generation is not supported.