As I mentioned in my earlier article, setting Cloud Operating Model is one of the first steps that an organization should start working on before moving first server/resource to Cloud.
Operating Model is a way in which People, Process and Platform should work together to ensure agility in technology development and business and achieve desired organizational goals.
Cloud operating model ensures that there exist right team and culture to take decisions, drive changes, evolve, innovate and grow while organization adopts Cloud.
An important point to note is that there is not just one team who could set up and execute ‘Cloud Operating Model’ principles and it is the responsibility of all the teams – IT Operations, IT Project and even Business. But there will definitely be a need of central governing body like Cloud Center of Excellence (Cloud CoE) who would ensure set up of Cloud Operating principles and drive that through. Some companies set up their own Cloud CoE team while some establish that in partnership with other service companies like Capgemini/Sogeti.
Following are important aspects that can act as guidelines while setting up and continuing with Cloud Operating Model:
- Define Goals: It is important to define business goals and map them with IT AND vice versa. Whether it is a new Value Chain establishment, faster release to market, more collaboration, new Channels to reach to market or increasing geographical footprint, Cloud can play a critical role into each one of these goals. Cloud gives the flexibility to explore new initiatives which can be dismantled easily if not working, set up new geographies pretty fast, implement DevOps pipelines, set up machine learning platforms at a rapid rate and cost-efficient rate. All the goals need to be associated with RoI (Return on Investment) analysis and tracked accordingly.
- Set up of Multi-Modal IT: It is important that traditional IT way of working and modern IT way of working exists not only across different applications but even for 1 large monolithic application as well. So, while part of a monolithic application continues to work in a traditional way, we break it part by part in microservices and deploy them in a modern way working closely with the existing remaining monolithic piece.
- Aim for Software Defined Everything: Defining everything through Software is essential. This brings automation, configuration management and agility to the environment.
- Set up of API Platform: Setup platform for APIs and Microservices to ensure ease of maintenance, scalability, reusability, and accountability.
- A balance between Portability Vs Flexibility: This argument is going to continue forever. While choosing one platform gives flexibility and agility to evolve, it does bind to the platform. While it is good to spread across, it also makes the management of multiple vendors/technologies complex. As long as that complexity is in proportion with the width & depth of portfolio and there is clear roadmap (at least for next 3-5 years) behind technology decisions, it should not be a problem.
- No compromise on Security: When in Cloud, security is shared responsibility. While cloud provider takes care of the security of physical data center, hardware, platform (in case of PaaS), users are responsible for non supported layers like OS (in case of IaaS), middleware and applications.
- Data: Data governance, storage, and usage are extremely important aspects apart from security and compliance. Cloud does provide robust and cost-efficient platform for data storage and provides on-demand large compute power to process data at scale with strong support for security and compliance. Data definitely is a relatively easier business case for Cloud.
- Collaboration: Culture of collaboration will encourage an exchange of ideas, sharing knowledge, the establishment of DevOps leading to a positive and innovative organization.
- Factory Set up: Set of factory teams for large tasks (like rehost/refactor of a large number of cloud migrations) containing a lot of repetitive and automation elements is very cost efficient and easier to maintain.
- Continuous Process: Cloud operating model is not one time process and it should continually evolve keeping pace with changing business situations.
Following diagram talks about different pieces and corresponding elements that would definitely need thought and potential implementation by teams responsible for Cloud Operating Model.
About Shailesh Gaikar
Shailesh is passionate about Cloud and IT transformation. He’s got an opportunity to work with clients across the Globe for cloud consulting, cloud assessment, cloud migration and cloud native development. Some of the key engagements are for cloud native development for Dow Dupont, migration for Thomson Reuters and Cloud Assessment for Gefco. Currently, he is the Head of Cloud and Microsoft for Sogeti Ireland and in the process of establishing cloud practice there.
More on Shailesh Gaikar.