Overview
Vambe’s e-commerce integration allows your AI assistants to recommend and discuss products with customers. Products are organized in Product Blocks (also called Prompt Blocks), which are containers that hold related products.Key Concepts
Product Blocks
- Container: A product block is a container that holds related products
- AI Context: AI assistants use product blocks to understand available products
- Organization: Organize products by category, brand, or use case
- Multiple Blocks: You can have multiple product blocks for different purposes
Product Structure
Each product has:- Basic Info: Name, description, price, currency
- Type: PRODUCT (physical/digital) or SERVICE
- External ID: Your unique identifier for tracking
- Metadata: Custom key-value pairs for additional data
- AI Description: Natural language description for AI understanding
Getting Started
Step 1: Get Your Product Block IDs
Before managing products, you need to know which product blocks are available:Step 2: Choose Your Product Management Strategy
Depending on your needs, choose the right endpoint:Endpoint | Use When | Behavior |
---|---|---|
Create | Adding new products or updating specific ones | Keeps existing products |
Upsert | Full catalog synchronization | Deletes products not in request |
Delete | Removing specific products | Deletes only specified products |
Complete Workflows
Workflow 1: Initial Product Catalog Setup
Workflow 2: Nightly Catalog Sync
Workflow 3: Add New Product
Workflow 4: Remove Discontinued Products
Best Practices
1. Writing Good Product Descriptions
The description field is crucial for AI understanding. Follow these guidelines: ✅ Good Description:- Use natural language that describes features and benefits
- Avoid URLs, prices (already in price field), or SKU codes
- Focus on what makes the product unique
- Include use cases and who it’s for
- Mention key features in conversational language
2. Use Meaningful External IDs
3. Organize with Metadata
4. Handle Large Catalogs
For catalogs with 1000+ products:Product Types
PRODUCT
Physical or digital items that customers purchase:- Physical products (clothes, electronics, etc.)
- Digital products (ebooks, software, etc.)
- Consumables (food, drinks, etc.)
SERVICE
Services or subscriptions:- Delivery services
- Consulting services
- Subscriptions
- Warranties
- Installation services
Currency Codes
Supported currencies include:- USD: US Dollar
- EUR: Euro
- GBP: British Pound
- CLP: Chilean Peso
- MXN: Mexican Peso
- COP: Colombian Peso
- ARS: Argentine Peso
- PEN: Peruvian Sol
- BRL: Brazilian Real
API Endpoints Reference
Product Management
- GET /api/public/product/product-blocks - List available product blocks
- POST /api/public/product/create/ - Add/update products (keeps others)
- POST /api/public/product/upsert/ - Full sync (deletes unlisted)
- POST /api/public/product/delete/ - Delete specific products
Order Information
- GET /api/public/order//product - Get products in an order
Common Scenarios
Scenario 1: E-commerce Platform Integration
Scenario 2: Real-time Updates
Scenario 3: Multi-category Management
Troubleshooting
Product Not Showing in AI Responses
- Check Description: Ensure description uses natural language
- Verify Block: Confirm product is in the correct block
- Check AI Configuration: Ensure assistant has access to product block
- Wait for Processing: Allow a few minutes for AI indexing
Products Being Deleted Unexpectedly
- Using Upsert?: Remember upsert deletes products not in request
- Solution: Use
create
endpoint for partial updates - Or: Include all products in upsert request
Metadata Not Appearing
- Check Format: Metadata must be array of
{key, value}
objects - Verify Keys: Keys should be simple strings
- Check Response: Verify metadata was saved
Related Resources
- Product Endpoints - Full API reference
- Create Pipeline - Set up AI assistant to use products
- Analytics - Track product recommendation performance
Next Steps
- Get Your Block IDs: Call GET /product-blocks
- Upload Products: Use create or upsert endpoint
- Configure AI Assistant: Link product block to your assistant
- Test: Chat with your AI and ask about products
- Monitor: Track which products are being recommended