Skip to main content

​
Brand Guardrails

Handing over your social media to AI requires trust. Our “Brand Guardrails” system is designed to ensure that trust is never broken.

​
Safety Layers

​
1. Content Moderation

We filter all generated text and imagery through enterprise-grade safety filters to prevent:
  • Offensive language.
  • Inappropriate imagery.
  • Competitor mentions.

​
2. “Never Show” List

During onboarding, you can define keywords or concepts to avoid.
  • Example: A vegan cafe can blacklist words like “meat”, “bacon”, “dairy”.
  • Example: A high-end salon can blacklist “cheap”, “discount”, “bargain”.

​
3. Visual Consistency

We digitally enforce your brand guidelines:
  • Color Locking: We ensure overlays match your hex codes.
  • Logo Safe Zones: We guarantee your logo is never obscured by captions or UI elements.

​
Human-in-the-Loop (Beta)

For select accounts or flagged content, we employ a “Human-in-the-Loop” system where a real human reviewer checks the video before it ever reaches your dashboard. This ensures an extra layer of quality control during our beta phase.

​
Feedback Mechanism

If a video doesn’t meet your standards, you can Reject it with a reason (e.g., “Music too loud”, “Wrong photo”).
  • This immediately stops the video from publishing.
  • It sends a signal to our AI to adjust future generations.
  • It triggers a regeneration so you don’t lose your content slot.

​
How to Configure Safety Settings

You can update your safety preferences at any time to refine what the AI produces.

​
Managing the “Never Show” List

  1. Navigate to Settings > Brand Guardrails.
  2. Scroll to the Negative Keywords section.
  3. Add Keywords: Type a word (e.g., “cheap”, “discount”) and press Enter.
  4. Save: The AI will now strictly avoid using these words in captions, text overlays, or voiceovers.

​
Reporting a Safety Issue

If a video slips through that you feel is inappropriate:
  1. Do NOT Publish.
  2. Click the Flag icon (instead of Reject).
  3. Select “Safety Concern” from the dropdown.
  4. Provide a brief detail.
    • Outcome: This creates a high-priority ticket for our Human Safety Team to review the generation logs and patch the filter.