Recommendations
Personalized recommendations, similar items, and trend detection.
Why This Exists
Helping users discover content they'll love requires understanding their preferences and finding similar content. Recommendations provide personalized suggestions based on user behavior, content similarity, and trends, so you can build discovery features without building recommendation algorithms yourself.
What It Is
Recommendations is WhizAI's capability for personalized content discovery. It uses vector embeddings, user behavior analysis, and collaborative filtering to suggest content users are likely to enjoy.
When to Use It
- ✅ Build personalized feeds ("For You" pages)
- ✅ Show "similar items" or "you might also like"
- ✅ Recommend content based on user behavior
- ✅ Surface trending content
- ✅ Increase engagement through discovery
Common use case: E-commerce "customers who bought this also bought" or content platform "recommended for you" feeds.
When NOT to Use It
- ❌ When you have very few users or items (recommendations need data to work well)
- ❌ For simple "most popular" lists (use analytics instead)
- ❌ When you need real-time personalization (recommendations update periodically, not instantly)
- ❌ For content that changes very frequently (recommendations work best with stable content)
Best practice: Start with similarity-based recommendations (simpler), then add personalization as you collect user behavior data.
API Endpoints
Similar Items
POST /v1/recommend/similar
Find items similar to a given item using vector embeddings.
curl -X POST https://api.whizur.com/v1/recommend/similar \
-H "X-API-Key: YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"itemId": "item_123",
"limit": 10
}'Personalized Feed
POST /v1/recommend/feed
Get a personalized content feed for a user.
curl -X POST https://api.whizur.com/v1/recommend/feed \
-H "X-API-Key: YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"userId": "user_123",
"limit": 20
}'Trend Detection
POST /v1/recommend/trends
Get trending items and trend analysis.
Billing & Usage
- Each recommendation request counts as one API call
- Usage is tracked per API key
API Endpoint Documentation
For detailed API endpoint documentation with request/response schemas, interactive testing, and examples, see the Interactive API Reference.
The API reference includes all recommendation endpoints: /v1/recommend/similar, /v1/recommend/feed, and /v1/recommend/trends.