Installation
Before installing the extension, please make sure to Backup your web directory and store database
You will require to install MageDelight Base Extension first. If it is not installed then please download it from https://www.magedelight.com/pub/extensions/magedelight-base.zip and unzip the package file into the root folder of your Magento 2 installation.
Root Directory is - "app/code/magedelight/#YourExtensionName#"
- Unzip the extension package file into the root folder of your Magento 2 installation.
- Connect to SSH console of your server:
- Navigate to the root folder of your Magento 2 setup
Run command as per below sequence,
php bin/magento setup:upgrade
php bin/magento setup:di:compile
php bin/magento setup:static-content:deploy
php bin/magento cache:clean
- log out from the backend and log in again
If you are using Magento 2.3 or later then you do not need to activate license. Extension will be activated with silent activation mechanism and does not require activation keys to be entered manually.
Backend Configuration (Admin side)
| Field | Description |
|---|---|
| Enabled | Select Yes to activate the AI Product Recommendation module or No to disable it. |
| Accuracy | Enter a value between 0 and 1 (e.g., 0.8). A lower value means higher accuracy. |
| Show Products | Enter how many products you want to display in recommendations (1 to 10). |
| Debug Mode | Select Yes to display the match percentage for each recommendation, or No to hide it. |
| Field | Description |
|---|---|
| ChromaDB Endpoint | Enter the URL of your ChromaDB server (e.g., as the vector database for product embeddings. |
| Server Endpoint | Enter the URL of your AI server endpoint (e.g., required for communication between Magento and the AI service. |
| Cache Lifetime (hours) | Define how long the AI product recommendations should remain cached before |
| API Key | Enter your API key (if required). This is used for authentication when connecting to external AI providers such as OpenAI or Gemini. |
| Delete ChromaDB Collections | Click Delete ChromaDB Collections to remove all existing ChromaDB collections for this store. This action resets stored embeddings. |
| Field | Description |
|---|---|
| Collection Name | The ChromaDB collection name for storing product embeddings (e.g., magento_products). |
| API Timeout (seconds) | Enter the timeout duration for ChromaDB API calls. Allowed range is 5 to 300 seconds. The default is 30 seconds. |
| Test Connection | Click Test Connection to verify the connection to the ChromaDB server. If successful, a message will display: “Connection successful! ChromaDB is reachable, and collection is ready.” |
Note: It is recommended not to change the Collection Name because the collection name is not dynamic. If you change it, you will need to update the collection name across all related files manually.
| Field | Description |
|---|---|
| Include Product Name | Select Yes to include the product name in the embedding calculation, or No to exclude it. |
| Include Product SKU | Select Yes to include the product SKU in the embedding calculation, or No to exclude it. |
| Include Description | Select Yes to include the full product description in the embedding calculation, or No to exclude it. |
| Include Short Description | Select Yes to include the product short description in the embedding calculation, or No to exclude it. |
| Include Meta Title | Select Yes to include the product meta title in the embedding calculation, or No to exclude it. |
| Include Meta Description | Select Yes to include the product meta description in the embedding calculation, or No to exclude it. |
| Custom Attributes | Enter a comma-separated list of additional product attributes you want to include in the embedding calculation (e.g., |
| Title | Description |
|---|---|
| Enable | Select Yes to activate browsing-pattern–based product recommendations on the storefront. Select No to disable this feature. |
| Display Position | Choose the storefront locations where the Browsing Patterns Recommendation block should appear. Available options include Home Page, Product Page, and Cart Page. Multiple selections are allowed. |
| Title | Enter the main display title for the Browsing Patterns Recommendations section. This title appears above the recommended products on the storefront. |
| Sub Title | Provide a sub-title for the recommendations block. This text appears below the main title and is used to describe the context of AI-powered browsing-pattern recommendations. |
| Sort Order | Define the sort order for the Browsing Patterns Recommendation block relative to other frontend components. A lower number displays the section earlier on the page. |
| Title | Description |
|---|---|
| Enable | Select Yes to activate past-purchase–based product recommendations. Select No to disable this feature. |
| Display Position | Choose the storefront pages where the Past Purchases Recommendation block should be displayed. Options include Home Page, Product Page, and Cart Page. Multiple selections are supported. |
| Title | Enter the main display title for the Past Purchases Recommendations section. This title appears above the recommended products on the storefront. |
| Sub Title | Provide a sub-title that appears below the main title. This text helps inform customers that the recommendations are based on their previous purchases. |
| Sort Order | Set the sort order number to determine the position of this recommendation block relative to other frontend sections. A lower value places the block higher on the page. |
| Title | Description |
|---|---|
| Enable | Select Yes to activate wishlist-based product recommendations on the storefront. Select No to disable this feature. |
| Display Position | Choose the storefront page locations where the Wishlist Recommendations block should appear. Available options include Home Page, Product Page, and Cart Page. You may select one or multiple positions. |
| Title | Enter the main display title for the Wishlist Recommendations section. This title is shown above the recommended products on the storefront. |
| Sub Title | Provide a sub-title for the recommendations block. This text appears below the main title and informs customers that the suggestions are based on items saved in their wishlist. |
| Sort Order | Specify the sort order for this recommendation block. A lower number places the block higher on the page relative to other sections. |
| Title | Description |
|---|---|
| Enable Unified Recommendations | Select Yes to activate the unified recommendations feature. This option combines multiple recommendation types—browsing patterns, past purchases, wishlist items, and current product similarity—into one consolidated block. Select No to disable the unified recommendation section. |
| Display Position | Choose the storefront pages where the Unified Recommendations block should be displayed. Available positions include Home Page, Product Page, and Cart Page. Multiple selections are supported. |
| Title | Enter the main display title for the Unified Recommendations section. This title appears above the consolidated recommendation block on the storefront. |
| Sub Title | Provide a sub-title to be shown under the main title. This helps describe the AI-powered unified recommendation functionality to customers. |
| Sort Order | Set the sort order value for determining the placement of the Unified Recommendations block relative to other frontend sections. A lower number displays the block earlier on the page. |
| Maximum Products | Specify the maximum number of products to display within the unified recommendation block. Accepted values range from 1 to 50. |
| Include Browsing Patterns | Select Yes to include recommendations based on customer browsing behavior in the unified group. Select No to exclude them. |
| Include Past Purchases | Select Yes to include recommendations derived from the customer’s previous purchases. Select No to exclude them. |
| Include Wishlist | Select Yes to include recommendations based on items saved to the customer’s wishlist. Select No to exclude them. |
| Include Current Product | Select Yes to include current product–based AI recommendations (vector similarity) in the unified group. Select No to exclude them. |
| Title | Description |
|---|---|
| Enable | Select Yes to activate current product–based AI recommendations. This feature displays products similar to the one the customer is currently viewing. Select No to disable this section. |
| Title | Enter the main display title for the Current Product Recommendations section. This title appears above the recommended products on the storefront. |
| Sub Title | Provide a sub-title to be shown under the main title. This text describes that the recommendations are generated by an AI model based on the similarity of the current product. |
| Sort Order | Define the sort order for this recommendation block. A lower number positions the block earlier on the storefront page relative to other content sections. |
Frontend Functionality
Storefront View – AI Recommendations
- Once the module is successfully configured, the storefront product page will display a new “Recommended for you” section powered by AI. This section automatically suggests products that are most relevant to the one being viewed.
- As shown in the image, when a customer views the Olivia 1/4 Zip Light Jacket product page, a recommendation block appears below the product details. In this example, it showcases four similar products — such as jackets, pullovers, and sweatshirts — that complement the current item.
- The recommendations are generated dynamically by the AI engine based on product embeddings and configured settings. This gives customers a personalized shopping experience, increases the chances of cross-sells and upsells, and helps them discover related products without leaving the page.
- In short, the frontend experience seamlessly integrates into the product page, showing an AI-powered “Recommended for you” slider with visually similar and relevant products.
Storefront View with 8 Product Recommendations (Example)
- When the admin sets the Accuracy value to 0.8 and the Show Products value to 8 in the backend configuration, the product page will display a broader set of recommended products while still maintaining a high level of relevance.
- As shown in the example image, when viewing the Olivia 1/4 Zip Light Jacket, the AI-powered recommendation block under the product details displays eight suggested items. These include similar jackets, pullovers, sweatshirts, and fleeces that are relevant to the customer’s browsing choice.
- This setup balances accuracy with variety — the AI recommendations are still precise enough to show related items, but the larger number of displayed products increases the chances for cross-selling and upselling.
Storefront View with Debug Mode Enabled
- When Debug Mode is enabled in the configuration, the storefront product recommendation section will also display a match percentage bar for each suggested product. This indicates how closely the recommended product matches the one currently being viewed.
- For example, in the product page shown, the Desiree Fitness Tee is displayed along with three AI-powered recommendations. Each suggested item (e.g., Aero Daily Fitness Tee, Diva Gym Tee, Gwyn Endurance Tee) includes a percentage score beneath it — such as 50%, 46%, or 44%. These percentages are visually represented with a progress bar, helping the admin or store manager understand the relevance of the recommendation.
- This feature is especially useful during testing and fine-tuning, as it provides transparency into the AI model’s decision-making and allows validation of whether the configured accuracy threshold is generating the desired results.
Storefront View – Recommendations on Cart Page
- If the Display Configuration is set to Cart Page or Product Page & Cart Page, the AI-powered product recommendations will also appear directly in the shopping cart.
- As shown in the example, below the cart items and summary, a “You might also like” section is displayed. This block contains AI-generated suggestions related to the products currently in the cart. The recommendations are based on product embeddings and configuration values defined by the store admin.
If Debug Mode is enabled, each recommended product will also display a match percentage bar, indicating how closely it relates to the cart item(s).
If Debug Mode is disabled, the recommendations will be displayed without percentages, keeping the customer-facing view cleaner.
- This setup enhances cross-selling opportunities during checkout, encouraging customers to add more relevant items before placing their order.
Based on Your Browsing Patterns
- The *Based on Your Browsing Patterns* section automatically displays AI-powered product suggestions based on the customer’s recent browsing behavior. Depending on the configuration, this recommendation block appears on the Home Page, Product Page, or Cart Page and showcases items that closely match the shopper’s interests. When Debug Mode is enabled, each product also displays a match-percentage bar to indicate similarity.
Past Purchases Product Recommendations
- The Based on Your Past Purchases section displays AI-driven product recommendations tailored to the customer’s previous buying history. It helps shoppers discover items that complement or closely relate to products they have purchased before, improving repeat engagement and cross-selling. When Debug Mode is enabled, each item also shows a match-percentage bar indicating how closely it aligns with past purchases.
Wishlist Product Recommendations
- The Based on Your Wishlist section showcases AI-powered product suggestions derived from items the customer has previously added to their wishlist. This helps shoppers rediscover saved interests and explore similar products they may love. When Debug Mode is enabled, each recommendation includes a match-percentage indicator showing how closely it aligns with the wishlist items.
Unified Product Recommendations
- The Unified Recommendations section combines multiple AI-driven recommendation sources—browsing patterns, past purchases, wishlist items, and current product similarity—into one consolidated product carousel. It delivers a richer, more personalized shopping experience by showing the most relevant items across all recommendation types. When Debug Mode is enabled, each product includes a match-percentage indicator that reflects overall relevance.
Current Product Recommendations
- The Current Product Recommendations section displays AI-generated suggestions for items that are similar to the product currently being viewed. These recommendations help customers explore related options without leaving the page, improving product discovery and conversion. When Debug Mode is enabled, each item includes a similarity match-percentage bar showing how closely it relates to the current product.
Thank you for choosing MageDelight!