Automate NSFW Detection in ImageKit with GPT-4o and n8n

Effortlessly Automate NSFW Detection in ImageKit with GPT-4o and n8n

Enhance your content moderation strategy with our advanced automation workflow designed to detect and manage NSFW content in real-time. This workflow seamlessly integrates ImageKit with the cutting-edge capabilities of OpenAI's GPT-4o run via OpenRouter. Ensure your digital assets are safe and appropriate by automatically unpublishing any flagged content before it reaches your audience.

What this workflow does

• Activation Initiated: The workflow kicks into action when ImageKit detects a new upload event (`file.created`).
• Image Processing: Downloads the newly uploaded image from its ImageKit URL in binary format.
• NSFW Evaluation: Leverages OpenRouter with the GPT-4o model to analyze the image, producing a JSON verdict including `is_nsfw`, `confidence`, and `reason`.
• Content Moderation: Automatically unpublishes images in ImageKit if the NSFW check returns true, thus keeping inappropriate visuals off your CDN.
• Conditional Workflow: If images are not flagged as NSFW, the process completes without alterations to the published content.

Use cases

• Ideal for content moderation teams requiring automated vetting of visual assets before exposure to a broad audience.
• Streamline content management for SaaS platforms that host user-generated content and wish to impose strict guideline adherence.
• Perfect solution for e-commerce stores seeking to prevent inappropriate content from affecting brand integrity.

Technical details

• Integrations: ImageKit for asset management, OpenRouter utilizing OpenAI GPT-4o for image analysis.
• Nodes utilized: if, sticky note, http request, n8nn8n-nodes-langchainagent, imagekitn8n-nodes-imagekitimagekit, n8nn8n-nodes-langchainlm chat open router.
• Prerequisites: Set up ImageKit credentials, register n8n webhook URL with ImageKit, and establish OpenRouter credentials for GPT-4o access.

With our streamlined automation process, you can confidently maintain an appropriate content library, allowing GPT-4o's precision and n8n's flexibility to ensure an efficient and protective workflow.