Skip to content

Why migrate to AWS to become a platform for IT department?


Becoming a Platform enterprise (see the book Platform Revolution: How Networked Markets are transforming the Economy) requires relying on an IT department that moves faster, leverages the most innovative technologies, and can scale in terms of transaction volume and internationalization.

Additionally, it involves opening the platform to customers and suppliers (while ensuring security and regulatory compliance), delivering a quality experience, optimizing costs per business transaction, and facilitating the understanding of platform usage.

So, where do you start?

Adopting a public cloud platform means building on foundations that will enable meeting these requirements by adopting best practices and a suitable operational model:

  • Building and deploying enterprise platform services faster, relying on managed services and automated workflows.
  • Leveraging emerging technologies by consuming new managed services (Data, AI/ML, IoT, and blockchain) to reduce the learning curve.
  • Drastically expanding usage by aligning capabilities with demand and deploying services in new geographical regions worldwide.
  • Opening and controlling platform usage using connectivity offered by public cloud providers while identifying and granting access to platform users based on their profile and usage context.
  • Safeguarding digital assets by implementing encryption, audit, and network security measures to facilitate regulatory compliance.
  • Ensuring users a quality experience regardless of location by leveraging a cache network available in all regions worldwide.
  • Measuring and optimizing costs per business transaction on the platform and adopting a digital frugality approach by adjusting resource capacities to user demand and profiles.
  • Consolidating platform usage data and applying Machine Learning algorithms to understand user behavior better, adapt existing services, and develop new ones.

How to shift your IT infrastructure from your existing data centers or managed service providers to a public cloud platform? Is executing a large-scale migration achievable for every company?

First of all, a large-scale migration must be viewed as a strategic program that contributes to the ambitions of the Platform enterprise, ensuring alignment with project objectives, resources, and timelines.

This alignment is usually achieved during the “Assess” phase, which clarifies the preparation steps required before commencing the large-scale migration.

In the “Mobilize” phase (preparation for the large-scale migration), four essential activities are carried out:

Building the Landing Zone

The “landing zone” construction structures the AWS environment intended to host your applications. This environment facilitates the platform’s opening while implementing controls on internet traffic through a SAS zone and identity and access management mechanisms via front-end applications or APIs (for both platform administrators and potentially end-users of its services).

In conjunction with the landing zone, identifying the security measures to be implemented (based on the company’s security policy and regulatory constraints) is done, as well as adapting the deployment and operation processes necessary to address the cloud platform.

Assessing the Application Portfolio

Reviewing the application portfolio determines the migration trajectory for each application and its migration planning. Several factors guide this trajectory:

  • The teams’ experience and capacity to acquire skills in managed services or serverless technologies and operational tools suitable for the cloud platform.
  • The available time to perform the migration is often dictated by hosting imperatives (data center release, expiration of a managed service contract) or business deadlines.
  • The positioning of applications in the Platform enterprise strategy:
  • Common applications that market SaaS applications can replace.
  • Core business applications that need to be optimized to ensure enterprise performance.
  • Applications that represent growth opportunities to offer differentiated platform services and relying on innovative architectures.

Defining the Target Operating Model

Defining the target operating model often involves establishing (or scaling) a cloud competence center and relationships between the cloud competence center, DevOps teams, and cross-functional teams such as networking, security, and tools. The plan for upskilling teams based on their target roles is also defined.

Additionally, Cloud governance is established with its managerial component (strategic alignment on the company’s cloud usage guidelines and the associated means) and its financial steering component.

Experimenting with Pilot Applications

Experimenting with the migration process on pilot applications validates:

  • The migration process and responsibilities.
  • The landing zone’s capacity to provide services expected by the platform applications.
  • The operational model and processes for applications hosted in the cloud.

Migrating at Scale and Optimizing

The last phase is the migration at the scale of applications and its optimization (“Migrate & Optimize”). It involves executing migrations per the plan developed in the previous phase.

This activity is carried out jointly by the DevOps Factories that understand the applications and engage with the business, the Migration Factory that brings expertise in migration methods and cloud architectures, and the Cloud Competence Center that provides expertise on the cloud platform. The Migration Factory also provides program consolidation, risk management, and cross-functional capitalization.

Migration of a Major E-Commerce Player

In 2020, Devoteam Revolve, in coordination with the client’s teams and AWS, prepared the migration of a significant European e-commerce player (11 million active customers, €4.3 billion in revenue in 2021), for whom any service interruption results in a loss of thousands of euros per second during peak activities.

Around 120 people were involved in this preparation, 7,000 physical and virtual servers across 2 data centers were inventoried, 350 applications were reviewed, and 250 were migrated. The large-scale migration was carried out in the first ten months of 2021, with the last two months reserved for the physical decommissioning of equipment and the broker takeover phase. The large-scale migration mobilized nearly 300 people to meet the deadline for freeing the two data centers, as per the initial planning. From October onwards, the client was able to achieve its highest daily revenue on the AWS platform.