As a user, you get the predictions in your apps through an application you’re already using. A problem software engineers face is deploying machine learning models into existing applications.
PostgresML aims to help users with a basic understanding of SQL build, deploy and maintain machine learning models. All in a productive environment with less risk and effort than other systems.
PostgresML makes machine learning simple by moving the code to your data rather than copying the data all over the place. It’s safe to say that it’s a great place to store your data. It also has an app that provides visibility into models & datasets in your database using the built-in dashboard.
- Scheduled training: There are some applications where data changes often. The retraining of models can is done on a schedule by PostgresML.
- Data explorer: This is another feature of PostgresML. A data explorer allows anyone to search for the dataset in production. It helps users browse useful tables and features to build effective models.
- Production deployment: Replicate production data in real-time to PostgresML.
- Aesthetically pleasing dashboard: PostgresML has a good looking dashboard and user friendly interface.
- Various algorithms: PostgresML has several algorithms; it can be regression or classification algorithms.
Functions Using SQL
- Training: PostgresML is a single call that can manage the different tasks of training. This function is at the heart of PostgresML.
- Prediction: The predict function is the key value proposition of PostgresML. It provides online predictions using the model that was actively deployed.
Example Use Case of PostgresML
PostgresML uses machine learning technique and models to conduct analysis for stock market price prediction.