Brand Safety

1. What should I do if a job submission is unsuccessful?

If a job submission fails, check the response for errors and re-submit the request with the correct parameters. Monitor webhook events for SAFETY_FETCH.FAILURE notifications. Check the Webhook Events page for more details.

2. How do I interpret the safety rating and content report details in the API response?

safety_rating indicates the overall risk level (e.g., LOW RISK, HIGH RISK). Scores that exceed the threshold are marked as HIGH RISK. The content report includes specific items and their associated risks, categorized and scored based on thresholds. The categories, their descriptions, and their threshold details can be found on this Airtable link.

3. What platforms are supported by the API for trust and safety analysis?

Supported platforms include Twitter, Instagram, LinkedIn, TikTok, Facebook and YouTube. More platforms are being added.

4. What media formats can the API process?

The API supports image, audio, video, and text formats.

5. How does the API calculate the risk scores?

The API analyzes media content against various parameters and compares the results to threshold values. Scores exceeding thresholds are marked as high risk.

6. What is the expected latency for all the operations?

Typically, responses to media URLs are instant. For profiles containing up to 10 items, you can expect output within 5 minutes. However, processing times may vary, particularly with video content which generally requires longer processing times.

7. Are timestamps included in the API response and reports?

Yes, timestamps are optionally (based on request) included in both, the API response and the downloadable reports. They provide contextual timing for when flagged content or events occurred within the media.