Tech leaders discuss how to optimise cost and security for cloud infrastructure
As IT moves to the cloud, so will IT costs. Saving money in the cloud requires organisations to assess their current spending and find opportunities for efficiencies.
About 60% of Indian enterprises’ revenue is expected to come from digital business models in the next three years, according to International Data Corporation’s India Digital Business Survey, 2023 – whether it is in terms of cost, performance, security, or ability to scale.
Cloud computing is projected to see a major upsurge with the multicloud strategy, which is revolutionising how businesses use the cloud. Oracle and YourStory brought together technology leaders to understand each of their respective company’s current status on the cloud journey.
The round-table discussion titled ‘Cloud Infrastructure for Scalable Performance and Cost Efficiency’ featured Shivam Maheshwari, Cofounder and CTO, Goals101; Aman Gupta, Director-Engineering, INDmoney; Sushan Rungta, CTO, Absolute; Hemant Mongia, Vice President - Technology, BharatPe; Diwas Sharma, Head - Technology, Fashinza; and Vipul Jain, AVP- Technology, BluSmart Mobility.
While one of the participants cited the use of Kubernetes to manage costs and optimise resources, and the significance of automation in cost management, another talked about the use of cloud cost explorers and tagging resources for cost attribution and budget allocation.
The challenges of managing compute, network, and storage in the cloud and the importance of aligning with industry best practices were also brought to the fore.
Security and compliance in cloud adoption
Highlighting the importance of having a centralised security solution, technology heads from various startups focused on using private subnets, unified firewalls, and prevention of distributed denial-of-service (DDoS) attacks. They also exchanged views on data encryption, using tokenisation for personally identifiable information (PII) data, using serverless architecture, and having a disaster recovery plan.
Serverless architectures, for example, allow businesses to operate without the need to manage physical servers, reducing the total cost of ownership. While eliminating the need to manage infrastructure, serverless computing can expose organisations to security risks if the infrastructure is not properly configured. Misconfigured serverless environments can lead to data breaches, unauthorised access or service disruptions.
Speaking about automation in cost management and the benefits of using cloud-native services, the participants concurred that it is crucial to be cloud-agnostic and be compatible with different clouds to avoid vendor lock-in. Through multicloud, businesses may move their workloads between platforms without being restricted by a single contract or pricing plan.
With Generative AI coming into mainstay, cloud costs are again going to soar. Experts agreed that it’s best to optimise their cloud costs while running a workload. As cost has to be measured in unit economics, it's crucial to have a great deal of automation in order to save on costs. Gen AI can be used in two segments – one is to serve end customers with prompt engineering to summarise data for the end user and secondly, to use it as a checker or as a reactive pipeline for various automations across use cases.
Cost optimisation and multicloud strategies
While the hybrid cloud integrates both public and private clouds, multicloud accesses many public cloud providers. This strategy lowers risk, reduces expenses, and increases flexibility by preventing businesses from depending just on one supplier. Experts agreed that multicloud is not just technology play but also cost play. Therefore, the need of the hour is to understand the pricing philosophy of every cloud.
“You can only improve what you can measure and if you have good metrics, you can start optimising not just the application, but also your deployments,” said one of the speakers.
If a company is scaling and policies are not optimised, it will incur a lot of extra cost without even being aware of it. Thus, it is important to experiment and fine-tune refinements in deployments and architecture applications.
Companies must formulate a strategy for managing egress and ingress costs and use a lakehouse architecture to enjoy its benefits. Businesses should build their code in a way that it can be deployed into multiple clouds. The core must remain the same because there would be less operating expenses to worry about when migrating.
The roundtable covered a range of important topics, including optimising cloud costs by implementing better monitoring, automation and cloud-native services. They also evaluated the role of GenAI in automating various business processes, such as summarising customer reviews and extracting insights from documents.
Panellists agreed that there is a need to investigate solutions to improve security and compliance in a multi-cloud environment, such as a centralised security framework. The discussion wrapped with a deep dive into GenAI capabilities, and how they need to be integrated into payment processing workflows to quickly analyse problems and provide summaries to banks.