Muvi Launches Four New Features for its Recommendation Engine – Alie

Muvi has just launched four brand new features for its AI-powered recommendation engine Alie. So far, Alie has stood out from the crowd being industry-agnostic, scalable and having various feature-rich solutions. And from now on, Alie customers will have more freedom to train Alie in their own ways! 

Here’s an Exclusive Insight to the Newest Features of Alie:

1. Training of Multiple Datasets

You can now add multiple datasets to train Alie for faster & improved recommendations. These datasets can be added from your local computer, using API/Webhooks integration or by simply copy-pasting a Javascript code to your website.

What Changes will this Feature Bring?

• Recommendations can be hooked in different places of a website or platform with different sets of algorithms and parameters
• Customers can train multiple data sets or same data set using different algorithms to hook the recommendations on different sections of their online platforms

Feature Highlights

• Action-based Recommendations
• Domain-specific Recommendations
• Enhanced End-user Experience
• Faster Implementation

A Use Case

By allowing you to train multiple data sets, Alie helps in offering action-based recommendations. These recommendations are based on the user’s history, items purchased, etc. Suppose, customer A has trained a specific set of data for his e-commerce store. End-user X purchased sports accessories on that e-commerce website, Alie will recommend other sports accessories to X in future.

For more information, check out our Multiple Data Set Training  Feature.

2. Selection of Data Type

A multi-domain recommendation system like Alie, needs to handle a diverse set of data types. This new feature lets you insert any type of data or parameters to the system to train or view the recommendation results.  All you need is to define the data type for each input parameter. Alie will then sync with the algorithm and train the data without any ambiguity.

How will this Feature Benefit the Customers?

• It will eliminate the data type ambiguity from the system
• More precise recommendations will be generated by training more transparent set of data
• The intelligence of Alie will be enhanced further by providing more clean and transparent data

Feature Highlights

• Improved Accuracy
• Industry-specific Recommendations

Example

A customer of Alie can set ‘string’ as data type for the parameter ‘user id’ and ‘double’ as data type for the parameter ‘price.’

You can know more about this newly added feature here.

3. Multiple Algorithms

Alie behaves differently for different algorithms and now the customers get to choose the algorithms according to their business needs.  We have listed various machine learning algorithms such as other user’s choice, most viewed, recently added and similar item recommendation, to name a few. You just have to log in to your dashboard, upload your own dataset and select your preferred algorithm. Alie will feed on the dataset and algorithm to generate the desired recommendations!

What Advantages will the Customers Get?

• Customers can choose the algorithms on the basis of their requirements
• In-house algorithms can be built for the customers to use
• More use cases can be added to the list

Feature Highlights

• Innovative Algorithms
• Alie’s In-House Algorithm
• Custom Algorithm

Example

X is an Alie customer who owns a OTT platform. 

Case 1: 

X chooses the algorithm based on other users’ choice.

Result to the End-users: User A has watched Batman, The Avengers, Thor. User B has watched Batman, Game of Thrones, Thor. So, Alie would recommend Batman, Thor, The Avengers, Game of Thrones to User C.

Case 2:

X chooses the algorithm based on the most viewed content.

Result to the End-Users: User A watched Interstellar which is of sci-fi genre. Now, Alie will recommend the most viewed movies or series in the sci-fi genre to A.

For more details, check out our Multiple Algorithms Feature.

4. Addition of Filters or Conditions

Good news for Alie customers who want to tailor recommendations for their end-users! Now you can configure input parameters to provide your users customized recommendations. This feature will have a customer-driven approach enabling you to add various filters or conditions for the recommendations generated to the end users. 

Feature Highlights

• Customize your Recommendations
• Use Logical Filters
• Industry Specific Recommendations

Example

Customer X is running an e-commerce store and doesn’t want to recommend the products below 3 stars to the shoppers. X can configure the parameter ‘rating’ by adding the condition that only products with rating greater than equal to 3 will be recommended to the shoppers. In the same way, various conditions can be added such as, equal to, not equal to, less than equal to, etc.

Have a look at our Customized Recommendations feature to get a more clear idea.

If you are yet to get Alie for your business, do it now!