Hybrid architecture & Machine Learning

  • Practice
  • Team

1 Architect, 1 Sr Data Engineer, 1 Sr Data Scientist

  • Technical environment
Successful PoC
on all criteria
Deployment at scale
of different Data Science Use Cases
Hybrid approach
to cloud use
  • The customer wanted an architecture that would enable it to use the computing power of the cloud to run Machine Learning models.

  • Securing data confidentiality was an important part of the mission.
  • Hybrid architecture (cloud and on-premise) to leverage the computing power of the cloud when training Machine Learning models, while preserving investments already made in on-premise.

  • Advice on the choice of the various components to be used on the Azure platform side, to ensure optimal use of resources and meet the level of complexity of Machine Learning scripts.

  • Data confidentiality was ensured through data desensitization and non-persistence in the cloud (with persistence time set at the customer's discretion).
  • The customer was able to validate a hybrid approach to the use of the cloud, validating these different requirements (security, compliance, confidentiality, etc.).

  • The solution met all the success criteria established for the PoC.

  • The customer is now ready to use the solution at scale for various Data Science use cases.