Reporting

Emily Reporting functionality helps developers and management

If you work with ML projects, you may be familiar with the lack of overview both for developers and managers. This is where Emily comes in handy.

Emily is the developer toolkit for helping machine learning engineers and data scientists in developing and deploying ML microservices.

Emily takes away the hassle right from setup of ML microservices and all the way through to deployment, as many of the usually manual steps are carried out automatically by Emily.

On top of this, it’s possible to obtain a complete overview of projects and experiments made with Emily, as the Emily Machine Learning templates come with an optional reporting functionality.

The reporting functionality will give you

Privacy

We know that models, experiments, and particularly data are incredibly sensitive. Therefore, Emily projects including the reporting functionality are setup either on your local machine, or on your remote server.

Collaboration on Enterprise level

With Emily reporting, you can gather insights from multiple Emily projects all in one place. Always having project status readily available makes it easier for teams to collaborate on projects. The report functionality also provides managers with an up-to-date status.

Full transparency and data version control

The reporting functionality creates snapshots of your code, data, model, experiment parameters, metrics, and results. It is therefore easy to link all these artifacts and even rollback to a specific experiment setup. This full transparency and version control is obtained by combining the best of MLflow and DVC. No setup is required; Emily will do the job for you.

Production maintenance

Always know which models are running in your environments. Emily reporting functionality provides a complete overview of models in your development, staging, and production environments. On top of that, Emily automates the re-deployment process, making swapping models really easy.

Dashboard

Emily reporting comes with an out-of-the-box MLflow dashboard, which you can access remotely from your browser. All your projects, experiment history, figures, and diagrams can be viewed providing a valuable overview.