Advanced Model Health
The Synerise Recommendation module enables users to create tailored recommendations across multiple channels using AI-based solutions, helping you reach your target audience with the most relevant items tailored to customer preferences.
Our Recommendation engine allows you to operate on several catalogs, which can contain different types of data, such as: products, articles, or even videos. Given that you can create multiple models for different types of objects and train them in a self-service manner, it is particularly important to control the state of these models.
To streamline the process of running and training your models, we have created a new feature that will provide you with additional information about their status.
How does the Advanced Model Health work?
From now on you have better control over the current status of your models! You can check if your model is trained, if it is in training, or if any errors have occurred during training. For more information on the different model statuses, please visit our Help Center.
If your business profile does not contain enough data to train a specific model, you will receive an error status, which will allow you to react faster and make any necessary corrections to the collected data in an efficient way, without wasting time on various doubts about what could have gone wrong and why the model is not working.
If you don't actively use certain models, they pause training after 10 days of non-use and turn off after 30 days, saving resources. You can turn them on manually at any time.
What are the main benefits of the Advanced Model Health?
- Better control over the status of models,
- Overview of training-related errors,
- Automatic deactivation of unused models,
- Clear error messages.
More details about Advanced Model Health can be found in our User Guide.
Take full advantage of the available knowledge about the state of your models, reacting immediately to any possible changes!