Service offering

[ 1 ] Machine learning environment

Polar TEP has implemented the MLflow platform to support machine learning activities.  MLflow is an open-source platform to manage all stages of the ML lifecycle, including experimentation, reproducibility, deployment, and a central model registry.

 

Once a processor has been trained in the machine learning environment, it can be ported to Polar TEP's processor execution environment for routine operational processing.

  • PTEP-ML
  • Polar TEP Machine Learning Environment
  • €1,500

    per month

[ 2 ] Processor Development Environment

Polar TEP provides two paths to support processor development:

  • A customized VM with all necessary software tools and libraries for development. Various configurations and languages are available upon request.
  • An Interactive Development Environment (IDE) based on Jupyter Notebooks.
  • PTEP-DE Int
  • Jupyter Notebooks
  • €550

    per month
  • PTEP-DE Small
  • 2 CPUs
  • 8 GB RAM
  • 32 GB SSD
  • Python Language
  • €550

    per month
  • PTEP-DE Medium
  • 24 CPUs
  • 832 GB RAM
  • 32128 GB SSD
  • Python Language
  • €650

    per month
  • PTEP-DE Large
  • 48 CPUs
  • 3264 GB RAM
  • 128256 GB SSD
  • Python Language
  • €750

    per month

[ 3 ] Processor Execution Environment

Polar TEP provides built-in processors, or users can develop their own (in the Polar TEP Development Environment or in the user's own environment) and implement them in the Polar TEP Processor Execution Environment through the use of Docker containers. A JSON definition file specifies user inputs and defaults for the processor that will be displayed in the processor's user interface. Processors can be used privately or shared with other users.

In the execution environment, processors have access to computation resources that scale dynamically with demand. Processors can be invoked manually through the user interface, or automatically for batch processing (processing of large data collections) or triggered processing (processing invoked as new data is received).

Two levels of access to the Processor Execution Environment are available and the first three months are free.

  • PTEP-EE Trial
  • Up to 500 CPU hours of processing time per month.
  • FREE

    for 3 months
  • PTEP-EE Standard
  • Up to 500 CPU hours of processing time per month.
  • €500

    per month
  • PTEP-EE Large
  • Up to 1000 CPU hours of processing time per month.
  • €1000

    per month

[ 4 ] Application Hosting Environment

For users who want to host their own applications on a VM within the Polar TEP environment, three sizes are available. Other configurations are available upon request. Ports can be opened for external communications.

  • PTEP-HE Small
  • 2 CPUs
  • 8 GB RAM
  • 32 GB SSD
  • Python Language
  • €550

    per month
  • PTEP-HE Medium
  • 24 CPUs
  • 832 GB RAM
  • 32128 GB SSD
  • Python Language
  • €650

    per month
  • PTEP-DE Large
  • 48 CPUs
  • 3264 GB RAM
  • 128256 GB SSD
  • Python Language
  • €750

    per month

[ 5 ] Additional Data Storage

Polar TEP Processor Development and Application Hosting Environments come with data storage. If additional storage is required, three types are available.

  • PTEP-DS SSD
  • Solid State Storage
  • €100/TB

    per month
  • PTEP-DS HDD
  • Magnetic Storage
  • €50/TB

    per month
  • PTEP-DS OS
  • Object Storage
  • €30/TB

    per month

[ 6 ] Processor Integration Support Service

Integration of processors into the Polar TEP Execution Environment is documented and examples are provided. If addition support is required, two levels are available.

  • PTEP-IS Small
  • Suitable for most processors (up to 5 days of effort)
  • €5,000

    One-Time Charge
  • PTEP-IS Large
  • Processor with complex processing chains and external data sources (up to 10 days of effort)
  • €9,000

    One-Time Charge

[ 7 ] Polar Data Management Support Service

The Polar TEP team includes the world-class expertise on all aspects of the management and use of polar data. They can provide advice to groups struggling with concepts and issues such as:

  • The polar data ecosystem
  • Ethically Open, FAIR, and CARE data principles
  • Federated search
  • Data publication, DOIs, and metadata
  • Vocabularies and semantics
  • Data interoperability standards and formats
  • Data processing tools and techniques such as machine learning and data cubes
  • PTEP-DM
  • Data Management Support Service
  • €1,000

    per day