Most enterprises will see the cloud as a source of innovation besides providing additional capacity. Gartner foresees 70% of tech workloads running in a cloud environment by 2028, up from 25% today.
While this enables IT systems to scale as needed, it can lead to cloud costs spiralling out of control. This is why Dr Werner Vogels, Amazon.com’s vice-president and chief technology officer, urges organisations to build cost-aware and sustainable IT architectures.
“We’re able to move fast [due to constraints removed by the cloud], but we have lost the art of architecting for cost or keeping cost in mind. We need to regain it. Cost is also a close proxy for sustainability, which is becoming a requirement — it can give you a good estimate of the amount of resources you use [since the cloud is based on the pay-per-use model],” he told delegates at the Amazon Web Services (AWS) re: Invent 2023 conference in Las Vegas last week.
A frugal systems architect
Vogels share the principles of becoming a frugal systems architect, which can be divided into three categories: design, measure and optimise. For design, he believes cost and sustainability must be part of non-functional requirements (NFR) when designing, developing and operating IT systems.
NFRs define how a system should perform and typically include accessibility, availability, scalability, maintainability, portability, security and compliance. By considering cost implications early and continuously, IT systems can be designed to balance features, time-to-market and efficiency.
See also: 80% of AI projects are projected to fail. Here's how it doesn't have to be this way
Organisations should also architect systems that align with the business model’s profit levers to achieve economies of scale as revenue permits. “Every design comes with trade-offs, so it is crucial to regularly re-evaluate technical and business trade-offs and invest in resources aligned to business needs,” he says.
Observability — or the ability to measure a system’s state and performance — is also key in ensuring cost is kept under control. Vogels gave the example of how two buildings with similar structures in Amsterdam were found to consume different amounts of power. The one that used less power had its electric meter near the front door, so occupants of the building were more aware of how much electricity they used.
He adds: “Observability [is crucial, and this can be achieved through] monitoring systems that help identify wasteful practices, streamline workflows, and strategically allocate resources to priorities.” For instance, the AWS Management Console myApplications enables organisations to monitor and manage their applications’ cost, health, security posture, and performance.
See also: Responsible AI starts with transparency
By having robust monitoring, organisations can tier their systems more effectively. They should decompose their applications into small blocks to categorise them based on how critical they are to the business and have switches or dials that allow granular control of those blocks to optimise the application for both cost and user experience. Vogels adds that IT teams should make those decisions with business teams because cost optimisation must be tied to business impact.
As organisations continue to rely on or use more of the cloud in the future, optimising the IT systems for cost must be an ongoing journey. He says: “Organisations must continually revisit their systems to find further areas of improvements. Continuously question what worked in the past. Revisit methods and tools despite previous successes. The savings we reap today fund innovation for tomorrow.”
Exemplary use of cloud
Bank of Ayudhya Public Company Limited (Krungsri) is among the organisations leveraging AWS to improve operational and cost efficiencies and enhance the customer experience.
Krungsri implemented a comprehensive data platform on AWS, which provides a holistic view of its retail customers and ensures efficient data storage and management. Powered by Amazon Simple Storage Service (a cloud object storage service), the platform securely and compliantly stores data from its customer base of 9.8 million, including account information and transactions from the bank’s departments. The bank also uses AWS data analytics services, like AWS Glue, to use data from its platform to develop new digital services such as personalised investing accounts.
Enhancing customer experience
Krungsri’s cloud data platform enables the bank to pursue machine learning (ML) projects that enhance customer experience. With Amazon SageMaker (a fully managed service to build, train, and deploy ML models), Krungsri can use ML to predict heavy traffic and cash demand in each of its ATM locations so that it can optimise delivery routes for vehicles transporting cash to more than 6,000 ATMs across Thailand. This led to fuel costs and transportation time savings while improving customer cash availability.
To stay ahead of the latest tech trends, click here for DigitalEdge Section
Since approximately 7.5% of the Thai adult population is either unbanked or underbanked, Krungsri plans to leverage ML to automatically assess loan risks for customers who do not own property or cannot provide financial documentation like pay slips.
It will also look at using AWS’s generative AI capabilities — including Amazon Bedrock, a fully managed service that makes foundational models available via an application programming interface — to drive productivity further and find new ways to serve customers.
“Our digital transformation into an AI-driven bank with AWS is helping us better serve our growing base of digitally savvy customers while also helping our underserved communities build financial resilience,” says Tul Roteseree, head of the Data and Analytics Division at Bank of Ayudhaya PCL.
He continues: “Our enterprise data platform on AWS brings all our subsidiaries together and helps them build digital and more inclusive financial services using data analytics and machine learning. AWS’s infrastructure’s reliability has also transformed how we bring solutions to the market to deliver accurate and up-to-date services for our customers. We’re particularly excited about collaborating with AWS to pursue generative AI projects to unlock efficiencies, optimise costs, and drive product innovation.”
