Home

AWS: A True Pioneer

Christofer Sandberg, CTO, Eyeota

Christofer Sandberg, CTO, Eyeota

Back in the 80s, there was a saying, “You cannot get fired for buying IBM.” The same was said in the 90s about buying Microsoft. I believe this quote is now applicable to AWS, as it extends the most optimal solution to clients for any specific problem in the IT industry. At Eyeota, we employed AWS in an effort to incorporate the hybrid IT infrastructure into our systems. This infrastructure is essentially a combination of in-house data center resources and cloud services, which results in the most efficient enterprise IT setup. By enabling a company to leverage the benefits of cloud’s agility, while maintaining the on-premise infrastructure, we have been able to benefit by adopting this integration. Eyeota has globally widespread customers, requiring the data center resources across various countries: in particular the United States, Europe, Latin America, Japan and Australia. This is to stay within close proximity to the end users, which has been possible because of the computing abilities that the cloud extends. Additionally, the cost factor plays a significant role in deciding whether or not to adopt AWS as your cloud service provider. A decade ago, committing to a large number of servers at once would require a 12-month payment upfront. With AWS now, we have the options of paying On-Demand, partially upfront or even on a monthly basis. We are currently saving almost 80 percent of our bill by combining on premise and AWS. Besides, the workload functionality at Eyeota is straightforward, the first component of which is data collection. This is primarily latency-sensitive, so the need to collect information from the end user by remaining close to them is vital. Additional functionality such as analytic data modeling performed on data sets is where the storage functions come into the picture, allowing the data to be fed back into our on-premise systems. Hence, the hybrid model where AWS is used for low latency workloads and on-premise for long term storage and analytics becomes the ideal infrastructure for us. Essentially, AWS can be leveraged to your advantage if the adequate amount of in-house knowledge exists, on how you can efficiently run it.

How Organizations Must Adapt to AWS Architecture

Oftentimes the query of adapting to AWS depends on the business itself.

AWS is new, modern and has a lot to offer. Even so, I believe it is history repeating itself, in regards to how software is currently acting as a service online

Strangely enough, various businesses benefit differently by employing the same architecture. The efficiency of this integration depends mostly on how much a company can benefit from it. For instance, if the efficiency has increased by a mere 5 percent post-integration, then the productivity of this employment is questioned. This scenario is particularly applicable in certain cases pertaining to capacity climbing. For instance, if the company has to scale up at the end of the month for billing reports, then the elasticity of the cloud can be utilized efficiently. A natural trend that is evidenced by larger companies is the incorporation of the cloud to avoid the hassle of performing these functions on their own; some may even choose to outsource. This, however, results in the loss of flexibility and ability to optimize the production margin. For companies that have end users geographically variant and have data collection as the first functionality, deploying cloud can result favorable; as having data centers all across the globe is very expensive.

Choosing the Right AWS Partner

Going with a partner/vendor to AWS would make sense for some companies; particularly for those that need a direct connection between their office and the AWS data center, for instance. The job of the partner in this scenario is simply to facilitate the functions that are out of the control of AWS. This scenario is most applicable to brick-and-mortar companies, wherein the IT part of it acts as a support function, rather than an integral part of the business. But then again, running services that are not necessarily part of the core business is often outsourced. Cloud on its own does not extend to clients, more than just the hardware level. But coupling with partners like Amazon or SAP is favorable in that they run it for you, in the form of Software as a Service (SaaS). Essentially, an AWS partner assists companies with their transitioning into the cloud more smoothly. For instance, consider an e-commerce website that is getting ready for its big Black Friday sale. This would equate to sudden spikes in site traffic, which is a hassle if you’re handling your own data center, requiring at least 4-6 months of preparation ahead of time. With the cloud, however, you can scale it out effectively. Even startup companies that primarily began with legacy systems can make use of the prototype-building capabilities offered by the cloud. This not only directs the company towards a more rapid development cycle but also minimizes the chances of errors.

Disruption Ahead

Trends continue to emerge as big companies keep on building their applications for internal use, but how they build them is going to vary. Instead of building them from a particular operating system like Linux or Windows, working on individual components and tying them together seems to gain traction. We are quickly approaching a stage where more focus can be driven towards business logic and the holistic fruition of the business. AWS is new, modern and has a lot to offer. Even so, I believe it is history repeating itself, in regards to how software is currently being developed by connecting various available frameworks and services. This is not far from what we had back in the early 80s with main-frames; it is being done on a more extensive, global level, but the concept remains unaltered.