Synthesise AI Integration Documentation
This document provides a comprehensive guide to the Synthesise AI integration within the AFFILIFY platform, detailing the architecture, API endpoints, data models, and implementation guidelines.
Synthesise AI Integration Documentation
AFFILIFY Multimedia Transformation
This document provides a comprehensive guide to the Synthesise AI integration within the AFFILIFY platform, detailing the architecture, API endpoints, data models, and implementation guidelines.
1. Overview
The Synthesise AI integration transforms AFFILIFY from a website builder into a Multi-Channel Affiliate Campaign Factory. Users can now:
- Generate professional video scripts using the "Hook, Story, Offer" framework
- Create AI-powered videos with realistic digital avatars
- Synthesize natural-sounding voiceovers
- Repurpose website content into social clips, podcasts, quote graphics, and ad variants
- A/B test personalized ad variants targeting different demographics
2. Architecture
2.1 Data Models
Core Multimedia Models
VideoAsset
- Stores generated video files and metadata
- Linked to Website and User documents
- Tracks video type (explainer, review, social, ad-variant)
- Includes thumbnail URLs and duration estimates
VideoScript
- Stores AI-generated video scripts
- Follows the "Hook, Story, Offer" framework
- Includes keyframes for scene composition
- Status tracking (draft, approved, in-production, completed)
VoiceAsset
- Stores voice configurations and clones
- Supports both synthesized and cloned voices
- Includes speed and pitch adjustments
- Enterprise feature: Voice cloning for brand consistency
AvatarConfig
- Stores avatar configurations
- Niche-specific avatars (tech, fitness, lifestyle, etc.)
- Tone selection (professional, casual, enthusiastic, authoritative)
- Default avatar management
ContentRepurposing
- Stores all repurposed content assets
- Includes social clips, podcast snippets, quote graphics, and ad variants
- Status tracking for async generation
- Performance metrics (CTR, conversions, impressions)
MediaSuggestions
- AI-generated recommendations for B-roll, thumbnails, and video styles
- Unsplash integration for stock footage
- Niche-specific suggestions
2.2 API Routes
Video Generation
POST /api/ai/generate-video
- Request body:
{ "websiteId": "string", "videoType": "explainer|review|social|ad-variant", "script": "string (optional)", "avatarId": "string", "voiceId": "string", "includeSubtitles": boolean, "resolution": "720p|1080p|4k" } - Response:
{ "success": boolean, "videoAssetId": "string", "videoUrl": "string", "thumbnailUrl": "string", "duration": number, "status": "ready|pending|failed" }
GET /api/ai/generate-video?videoAssetId=string
- Retrieves video generation status and details
Content Repurposing
POST /api/ai/repurpose-content
- Request body:
{ "websiteId": "string", "sourceType": "website|video|article", "includeTypes": ["social-clips", "podcast-snippet", "quote-graphics", "ad-variants"], "targetPlatforms": ["tiktok", "instagram", "youtube-shorts"], "demographics": ["young-professionals", "parents", "enthusiasts"] } - Response:
{ "success": boolean, "repurposingId": "string", "assets": { "socialClips": number, "podcastSnippet": boolean, "quoteGraphics": number, "adVariants": number }, "status": "completed|processing|failed" }
GET /api/ai/repurpose-content?repurposingId=string
- Retrieves repurposing status and asset details
Multimedia Management
PUT /api/websites/[websiteId]/multimedia
- Updates website multimedia data (scripts, voice preferences, avatars)
GET /api/websites/[websiteId]/multimedia
- Retrieves multimedia data for a website
3. Frontend Components
3.1 MediaStudio Component
Located in src/components/MediaStudio.tsx
Features:
- Script editor with Hook, Story, Offer sections
- Real-time word count and duration estimation
- Avatar selector with visual previews
- Voice selector with audio samples
- Scene composition visualization
- Low-fidelity video preview
- Save and generate functionality
Props:
interface MediaStudioProps {
websiteId: string;
initialScript?: {
hook: string;
story: string;
offer: string;
fullScript: string;
};
onSave?: (script: any) => void;
}
3.2 Dashboard Pages
Media Studio Page (/dashboard/media-studio)
- Full-featured video script creation interface
- Integrated MediaStudio component
- Quick action buttons for next steps
- Script writing tips and best practices
Content Repurposing Page (/dashboard/content-repurposing)
- Website selection for repurposing
- Asset type selection (social clips, podcast, quotes, ads)
- Generated assets management
- Performance tracking
My Websites Enhanced (/dashboard/my-websites-enhanced)
- Enhanced website cards with multimedia status
- Video asset count display
- Quick access to Media Studio and Repurposing
- Conversion rate analytics
4. AI Integration
4.1 Gemini 2.5 Pro Integration
The platform uses Google Gemini 2.5 Pro for:
-
Video Script Generation
- Analyzes website data and product information
- Generates Hook, Story, Offer structure
- Optimizes for video duration and pacing
- Includes [VISUAL CUE] markers for editors
-
Social Media Content
- Creates platform-specific clips (TikTok, Instagram, YouTube)
- Generates captions and hashtags
- Optimizes for engagement
-
Podcast Snippet Generation
- Creates conversational, engaging audio content
- Includes intro, main story, and outro
- Natural pacing and tone
-
Quote Graphics
- Generates shareable, emotionally resonant quotes
- Highlights specific benefits
- Includes subtle CTAs
-
Ad Variant Generation
- Creates demographic-specific ad scripts
- Tailors value propositions to audience needs
- Suggests avatar tone and voice style
4.2 Synthesise AI Integration
The platform integrates with Synthesise AI for:
-
Video Generation
- Converts scripts to realistic video with digital avatars
- Supports multiple avatar styles and tones
- Includes automatic subtitle generation
-
Voice Synthesis
- Natural-sounding voiceovers from scripts
- Multiple voice options and languages
- Speed and pitch adjustments
-
Voice Cloning (Enterprise)
- Custom voice clones for brand consistency
- Maintains voice characteristics across all videos
5. Implementation Guide
5.1 Setup
-
Environment Variables
GEMINI_API_KEY=your_gemini_api_key SYNTHESISE_AI_API_KEY=your_synthesise_ai_key SYNTHESISE_AI_API_URL=https://api.synthesise.ai/v1 MONGODB_URI=your_mongodb_connection_string JWT_SECRET=your_jwt_secret -
Database Setup
- Ensure MongoDB is running
- Models will auto-create collections on first use
-
Dependencies
@google/generative-ai: Gemini API clientmongoose: MongoDB ODMjsonwebtoken: Authentication
5.2 Usage Flow
- User creates/selects a website
- User navigates to Media Studio
- System fetches website data
- User edits video script (Hook, Story, Offer)
- User selects avatar and voice
- User clicks "Generate Video"
- System calls Gemini to create script (if not provided)
- System calls Synthesise AI to generate video
- Video asset saved to database
- User can repurpose content from generated video
5.3 Repurposing Workflow
- User selects website
- User chooses asset types to generate
- User clicks "Repurpose Content"
- System generates:
- Social clips (3x 30-second videos)
- Podcast snippet (2-3 minutes)
- Quote graphics (5x shareable quotes)
- Ad variants (3x demographic-specific ads)
- All assets saved and linked to website
- User can download/share assets
6. Plan Limits
Basic Plan
- Video generation: 5/month
- Voice synthesis: ✓
- Voice cloning: ✗
- Content repurposing: ✓
- Ad variants: 2
Pro Plan
- Video generation: 20/month
- Voice synthesis: ✓
- Voice cloning: ✗
- Content repurposing: ✓
- Ad variants: 5
Enterprise Plan
- Video generation: Unlimited
- Voice synthesis: ✓
- Voice cloning: ✓
- Content repurposing: ✓
- Ad variants: Unlimited
7. Error Handling
Common Errors
"Synthesise AI API key not configured"
- Solution: Set
SYNTHESISE_AI_API_KEYin environment variables - Fallback: Uses placeholder video URLs
"Failed to generate video script"
- Cause: Gemini API error or invalid website data
- Solution: Check website data completeness, verify API key
"Video asset not found"
- Cause: Invalid videoAssetId or user doesn't own asset
- Solution: Verify asset ID and user authentication
8. Performance Optimization
Caching
- Video scripts cached in database
- Avatar and voice configurations cached client-side
- Repurposing results cached for 24 hours
Async Processing
- Video generation runs asynchronously
- Status can be checked via GET endpoints
- Webhook support for completion notifications
CDN Integration
- Video files served via Netlify CDN
- Thumbnail optimization for web
- Automatic format conversion
9. Security Considerations
-
Authentication
- All endpoints require JWT token
- Token validated before processing
-
Authorization
- Users can only access their own websites and assets
- Database queries include userId filter
-
API Key Management
- API keys stored in environment variables
- Never exposed in client-side code
- Rotate keys regularly
-
Data Privacy
- User data encrypted in transit (HTTPS)
- Database connections use SSL
- No personal data shared with third-party APIs
10. Monitoring & Analytics
Tracked Metrics
- Video generation success rate
- Average video generation time
- Repurposing asset counts
- User engagement with generated content
- A/B test performance (CTR, conversions)
Logging
- All API calls logged with timestamps
- Error logs include stack traces
- Performance metrics tracked
11. Future Enhancements
-
Real-time Collaboration
- Multiple users editing scripts simultaneously
- Comment and feedback system
-
Advanced Analytics
- Video performance tracking
- Audience sentiment analysis
- Recommendation engine for optimization
-
Custom Avatars
- User-uploaded avatar support
- Avatar training from video samples
-
Workflow Automation
- Scheduled content generation
- Automatic repurposing on new website creation
- Multi-language support
-
Integration Marketplace
- Native integrations with social platforms
- Direct publishing to TikTok, Instagram, YouTube
- Email marketing platform integrations
12. Support & Troubleshooting
Common Issues
Videos not generating
- Check Synthesise AI API key
- Verify website data is complete
- Check MongoDB connection
Scripts not saving
- Verify JWT token is valid
- Check database permissions
- Ensure website ID is correct
Slow performance
- Check network latency to Synthesise AI
- Verify MongoDB indexes are created
- Consider caching frequently accessed assets
Getting Help
- Check logs:
console.logoutput in server logs - Test endpoints with Postman
- Verify environment variables are set correctly
13. Code Examples
Generating a Video Script
const response = await fetch('/api/ai/generate-video', {
method: 'POST',
headers: {
'Content-Type': 'application/json',
},
body: JSON.stringify({
websiteId: '507f1f77bcf86cd799439011',
videoType: 'explainer',
avatarId: 'professional-avatar',
voiceId: 'warm-voice',
}),
});
const data = await response.json();
console.log('Video generated:', data.videoUrl);
Repurposing Content
const response = await fetch('/api/ai/repurpose-content', {
method: 'POST',
headers: {
'Content-Type': 'application/json',
},
body: JSON.stringify({
websiteId: '507f1f77bcf86cd799439011',
sourceType: 'website',
includeTypes: ['social-clips', 'podcast-snippet', 'quote-graphics', 'ad-variants'],
demographics: ['young-professionals', 'parents', 'enthusiasts'],
}),
});
const data = await response.json();
console.log('Content repurposed:', data.repurposingId);
14. Conclusion
The Synthesise AI integration represents the next evolution of AFFILIFY, transforming it into a comprehensive multimedia marketing platform. By combining the power of deep-scrape website generation with AI-powered multimedia synthesis, AFFILIFY now enables affiliates to create complete, multi-channel marketing campaigns from a single product link.
This integration maintains the "Rolex" philosophy of quality, professionalism, and understated elegance, ensuring that every generated asset meets institutional standards.
Last Updated: January 23, 2026 Version: 1.0.0 Status: Production Ready
Related Documents
Design Document: BharatSeva AI
BharatSeva AI is a multi-agent orchestration system built on AWS using Amazon Bedrock Agents with Claude 3.5 Sonnet as the foundation model. The system deploys 10 AI agents (1 Master Orchestrator + 9 Specialist Agents) to assist India's informal sector workers in navigating government schemes across three domains: PM Vishwakarma (artisan credit), PMFBY (crop insurance), and BOCW (construction worker welfare).
OpenClaw Enterprise Transformation Plan
Transform OpenClaw from a single-user personal AI assistant into a **dual-mode platform** that is simultaneously:
Qwen Image and Edit: Open-sourcing and Local GGUF Generations with Lightning
Daniel Sandner, for article on https://sandner.art/
PolyAgent: Production-Grade Agentic Platform Architecture
PolyAgent is a **production-ready enterprise multi-agent AI platform** that combines industry best practices with a refined three-layer architecture optimized for reliability, security, and performance. The platform leverages **Go for orchestration** (Temporal workflows), **Python for AI intelligence** (LLM services), and **Rust for secure execution** (WASI sandbox), delivering sub-second response times with comprehensive observability.