At Merapar, we are proud to be a key AWS partner with the Media & Entertainment competency, showcasing our expertise in cloud-native solutions. In today's fast-evolving digital world, a strong cloud strategy is essential—whether through cloud-native engineering, on-premise systems, or hybrid solutions. Understanding the right approach requires experienced partners who can build and guide these journeys. In this interview, we speak with Kevin Savina, principal Partner Lead at AWS, about how these elements help M&E companies stay agile, scalable, and resilient while optimizing costs and reducing risks in their cloud journey.
Throughout my career, I've been deeply involved in helping media companies optimize their business and operations through various technology transitions— from Analog to Digital, Standard Definition (SD) to Ultra High Definition (UHD), and from hardware-based solutions to IT and software-driven approaches. Today, the shift to cloud and the emergence of AI/ML technologies represent the next phase in this evolution. These changes are compounded by rapid shifts in audience behavior and expectations, such as the rise of digital and social viewing platforms and the demand for smoother user experiences and more content.
I see the cloud as an invaluable tool for media companies to navigate these changes. It’s not just a shift in technology that allows for more granular operations and scalability; it also represents a paradigm shift that can drive a cultural and mindset change within companies. This enables unprecedented business agility and adaptability—key factors for staying competitive in an ever-evolving industry.
In the area of media, if we consider the entire chain, hybrid systems will be around for ever. It’s obvious that, taking the example of coverage of a sporting event, cameras for example will always have a physical component. So, in the world of Media & Entertainment, the question becomes more: how do we optimize these hybrid systems, and smartly grow the “cloud components” in them to ensure alignment with the business objectives (that can be cost, scale, agility, resilience or any other relevant business KPI)?
I have observed it’s often complicated to have an informed discussion about IT costs in organisations, because there are a lot of hidden / unseen / un-measured aspects to these costs. Instead of getting stuck in comparative costing, I have observed that working back from business KPIs and Metrics from the operations helps focus all stakeholders around a business outcome and then allows to then make the right infrastructure decisions.
I see the main reason for the delays in cloud migrations related to skills. Enterprises still don’t have all the cloud skills they need. This affect both planning and execution. Companies don’t have the internal cloud knowledge to evaluate which of the projects make sense to start with. Nor do they have the skills to efficiently run the migration projects they have started. The strain on these limited skilled resources increases time-to-value of these cloud migrations, thus delaying overall modernization strategies.
To convince organizations of this kind to make efforts, a multi-step approach is needed. First, convince leadership to give the right level of attention to the potential business value of cloud migrations Secondly, pick the right projects or use cases to start with and lastly build the right teams (mixing internal and sometimes external support) to create early success and show value.
The benefits of cloud computing can manifest themselves differently for different typologies of use cases. It’s important to have that in mind to have reasonable expectations of the benefits that can be achieved through cloud. Agility and scalability are obvious benefits of cloud, but systems need to be architected in the correct way to be able to get this value at a business level.
Cost reduction is a more complicated discussion. Depending on how cost is measured (what is in-scope and what is out-of-scope) the results can be biased. Cloud can be a contributor to cost reductions, but I feel a better way of framing the question is that with Cloud costs are more correlated to operational activity (and if the operational activity is correlated to business value, then cloud costs become correlated to business value)
Not all companies benefit equally from cloud migrations. A condition to benefit maximally from Cloud migrations is that the architectures are optimized for cloud. This is almost never achieved through lift & shift and requires a process of continuous improvement to continuously adapt the architecture to fit the business needs. The more efficient that continuous improvement process is, the more value can be delivered by cloud operations.
One of the cloud's greatest strengths is its agility and speed of execution, with fewer infrastructure constraints than traditional on-premises IT. For rapidly evolving technologies like AI, ML, and Generative AI, hyperscalers like AWS facilitate scaling. This advantage is further enhanced by AWS's managed services, such as SageMaker and Bedrock, which help lower the cost and complexity of adopting AI/ML technologies.
Architectural design is very important to be able to deliver optimal value of a system. This means one should walk back from the business needs (functional and non-functional) in order to define the architecture. It is not automatic that IT systems should necessarily reside in the cloud primarily. But experience shows that after analysis, it is often the case that Cloud-based architectures often prove to be better adapted to the business needs.
One of the biggest factors slowing down the ability of enterprise customers to get maximum value from cloud is the lack of skills. It takes time and practice to build up skills and experience on cloud. External partners, who have been working with AWS, invest in the skills on a continuous basis and have the experience of running a variety of projects are very important to accelerate the time to value for customers. They help accelerate projects, avoid traditional pitfalls, and reduce the number of iterations to reach optimized architectures.
So, My advice would be: look at your business objectives, and identify teams that can help deliver them fast. Understand what the skill level of your own team is, and don’t hesitate to enhance these teams with external experienced resources where it makes sense .