Using VAST Ingestion for HTML5 Ads

The VAST ingestion feature allows API customers to submit HTML5-based ads for video predictions without manually converting them into video formats. This simplifies the ingestion of the assets and provides correct predictions through our API.

Why Use VAST Ingestion?

Many Neurons AI partners use HTML5 players for their ads instead of traditional video formats. The Neurons API now supports VAST ingestion, which automatically converts HTML5 ads into videos. This enables:

✅ Seamless compatibility with AI predictions.
✅ No need for manual conversion of HTML5 to video.

How It Works?

The VAST ingestion process follows these steps:

Step 1: Submit Your VAST File & Demo URL

• Upload a VAST-compliant XML file (URL required).
• Provide a render URL, which is the page where the VAST-based ad is displayed.
• Optionally specify the device type ("desktop" by default or "mobile" for mobile-only ads).

Step 2: Validation Check

• The API ensures the VAST file and render URL are correctly formatted.
• If there are any errors, you will need to adjust the file or URL before proceeding.

Step 3: Automatic Conversion

• A headless browser loads the HTML5 ad from the render URL.
• The system records the ad based on the size and duration in the VAST file.
• The API returns a media ID, which is used to track the recorded video.

Step 4: AI Processing & Results

• The converted video enters the standard Neurons video pipeline.

• AI predictions generate insights such as heatmaps, attention metrics, and engagement data.
• Webhooks notify users when processing is complete. See details in the document. 

Important Considerations

🔹 Limited VAST Support – The system only reads basic details like ad size and duration. Other attributes are ignored.
🔹 No Player Validation – The system assumes the render URL functions correctly; users must ensure their player works.
🔹 Fixed Viewport Support – Only desktop and mobile viewports are supported.
🔹 Customer Responsibility – Users must adapt their VAST file for any custom triggers (e.g., interactivity).
🔹 Error Handling – If the request is incorrect, you may receive errors like 400 (Bad Request) or 404 (Not Found).

Once the VAST-based video is created, you can proceed to AI prediction analysis by using the media ID to retrieve results.