FansFrame: Self-Serve UGC Campaign Platform with AI Moderation
The Challenge
Brands running user-generated content (UGC) campaigns faced several pain points:
- Manual content moderation was slow, expensive, and inconsistent
- No self-serve platform existed for brands to launch campaigns independently
- UGC submissions needed to be stored, organized, and rights-managed at scale
- Brands needed real-time dashboards to track campaign performance
- Content moderation had to handle images, videos, and text across multiple languages
- Storage costs were spiraling with large media files
Existing solutions required enterprise contracts and dedicated account managers - pricing out smaller brands and agencies.
The Solution
We built FansFrame as a self-serve SaaS platform where brands can launch, manage, and moderate UGC campaigns:
Core Features
1. Self-Serve Campaign Builder
- Brands create campaigns with custom submission forms, rules, and branding
- Embeddable widgets for websites and social media landing pages
- QR code generation for physical activations (events, packaging)
- Multi-campaign management from a single dashboard
2. AI-Powered Content Moderation
- Dual-model approach: Claude for nuanced content analysis, GPT-4 Vision for image moderation
- Automated flagging of inappropriate content, brand safety violations, and spam
- Configurable moderation rules per campaign (strictness levels, custom blocklists)
- Human review queue for edge cases with AI confidence scoring
3. Cloudflare R2 Media Storage
- Cost-effective storage for high-volume image and video uploads
- Automatic format optimization and thumbnail generation
- CDN delivery for fast global access
- Rights management metadata attached to every asset
4. Real-Time Campaign Analytics
- Live submission tracking and engagement metrics
- Moderation pipeline visibility (pending, approved, rejected)
- Export capabilities for approved content packages
Development Process
- Weeks 1-2: Platform architecture, auth, campaign CRUD
- Weeks 3-5: Submission pipeline, R2 integration, media processing
- Weeks 6-8: AI moderation system, queue workers, analytics
- Weeks 9-10: Beta testing with 3 launch partners
Technologies & Tools
The Results
AI moderation accuracy across image and text content
Faster content moderation vs manual review
Reduction in storage costs with R2 vs S3
Time for a brand to launch a new campaign
Technical Deep Dive
AI Moderation Pipeline
FansFrame uses a dual-model moderation approach for maximum accuracy:
- Image Analysis: GPT-4 Vision scans for inappropriate content, brand safety violations, and quality standards
- Text Analysis: Claude evaluates captions and metadata for sentiment, spam, and policy compliance
- Confidence Scoring: Each submission gets a confidence score - high-confidence passes go straight through, low-confidence items enter the human review queue
- Learning Loop: Moderator decisions on edge cases refine the AI rules over time
Queue Architecture
The platform handles submission spikes (common during live events) through Laravel's queue system with dedicated workers for:
- Media upload processing and optimization
- AI moderation requests (rate-limited per provider)
- Notification dispatch (email, webhook)
- Analytics aggregation
Cost Optimization with Cloudflare R2
Moving from S3 to Cloudflare R2 eliminated egress fees entirely - a significant saving for a media-heavy platform where approved content is frequently downloaded by brands.
"FansFrame lets us run UGC campaigns without a massive team. The AI moderation catches things our human reviewers missed, and the self-serve setup means we can launch a campaign in minutes instead of weeks."
Want Similar Results for Your Business?
We specialize in delivering measurable improvements through custom development and AI integration. Let's discuss how we can help you achieve your goals.
More Case Studies
Explore other success stories
