Modules

These page contains the main modules of frontier that will be leveraged by the users. Each module has it's own separate functionality.

• Reports & Visualization

• Processing & Pipelines

• Data

• Model Registry & Node Instance

Reports & Visualization

This module contains all of the chart functions required to generate charts from data contained in frontier. The data and visualizations that have been configured can be turned into reports that can be shared with potential members.

Processing & Pipelines

This module contains various pipelines, each with its own set of rules. Frontier also allows you to add and use custom functions. These pipelines are a key feature of Frontier, allowing users to retrieve data from databases, train models on them, push data back into databases, vectorize, and much more.

Data

This module holds all of the data that the frontier system has to offer. It allows users to preview data, investigate and perform EDA, as well as check the database connection, specified queries, and so on.

Model Registry

This module holds all of the performance metrics for the models that were trained with the frontier system. Additionally, the trained model can be supplied through API, and drift information can be obtained.

Node Instance

This module displays all of the project's jupyter notebooks that have been launched. The user can also choose to kill the resources in order to free up the previously used resource.