AI Video Analysis System for Social Media Content

About Client

CarVertical is a global automotive data company operating across multiple markets and channels. The team works with large volumes of video content published on platforms such as YouTube, Instagram, TikTok, and Facebook, where brand mentions, partner codes, and content compliance need to be monitored at scale.

Clients challenge

Manual partner video review didn’t scale as each video had to be checked manually to identify brand mentions, partner codes, and relevant context, which became inefficient as content volume grew.

Solution

We designed and implemented a custom AI automation system tailored for carVertical’s content workflows.

1. Unified video ingestion
Each provided video URL is automatically downloaded and processed, regardless of the platform.

2. Platform-specific scraping at scale
Using Apify, we configured and tested dedicated scrapers for YouTube, Instagram, TikTok, and Facebook.

3. Gemini for video and audio analysis
Google Gemini API analyzes video, audio, and descriptions to detect brand mentions, partner codes, and content signals based on carVertical’s evaluation structure.

4. Centralized storage and reporting
Videos are stored in Google Cloud Storage, while structured analysis results are saved to a database / Google Sheets and synced automatically via App Script.

The results

• Automated analysis across 4 major social platforms

• Near-zero manual review effort

• Consistent, structured AI evaluations per video

• Scalable system ready for additional platforms

• Delivered in 30 days with a clear, extensible architecture