The Data Science and Machine Learning (DSML) Studio in Zoho Analytics should facilitate the development of customized models.
For example, the machine learning models companies build with it can be used to analyze and predict customer turnover. This is conveniently combined with Zoho’s existing solutions, which generate a lot of data about customers because they are designed to support companies in sales and marketing. A lot of customer data is generated in these departments.
DSML Studio includes the no-code assistant AutoML to analyze the data. This assistant helps users train, test, compare, deploy, and manage models. AutoML provides features for feature engineering, hyperparameter tuning, and comprehensive model analysis.
The Code Studio is also available for more advanced development work. This integrated Python coding environment allows users to create custom machine learning models and import existing Python models or libraries for subsequent execution within the platform.
Data management
With the Analytics update, Zoho also introduces new data management capabilities that should improve the accuracy and applicability of models. For example, the platform now offers Stream Analytics and 25 new data connectors, bringing the total number of connectors to more than 500.
Zoho Analytics users can now also set up and manage complex ETL pipelines. The extract, transform, load (ETL) principle requires several steps. It starts with creating end-to-end data pipelines in the visual builder, after which users can develop custom transformations and ML models via the Python Code Studio. Work can also be done through the Zoho copilot Ask Zia as part of the transformation process. Finally, there is access to an automatic version control feature and a new Sandbox environment, and pipelines can be orchestrated using Zoho Flow.