Infosys TechCompass

Platform as a service

Trend 1: Container offerings evolve to build enterprise-ready solutions

Using containers for building scalable and portable microservices has become the de facto standard across all digital platforms. Kubernetes emerged as a standard for orchestration and management of container deployments. Every hyperscaler already had its version of managed Kubernetes, allowing customers to deploy containers at scale. This past year has seen further evolution of these offerings to make them enterprise-ready and a complete stack for building applications. As more applications are deployed onto Kubernetes clusters, more sophisticated requirements for connectivity, observability, and security lead to extended offerings of service mesh components such as Istio, Open Service Mesh, and Dapr.

Deployment pipelines are more automated, with secrets management and container image scanning built in. Serverless versions of container services such as AWS ECS Fargate, Azure Container Apps, and Google Cloud Run have been added to run one-of batch kinds of workloads. Event-driven pod execution and on-demand scaling have been added with the integration of KEDA, Knative, and Virtual Kubelet frameworks. Ingress services have been integrated with the cloud provider's load balancer (e.g., AWS Elastic Load Balancing) and gateway services (e.g., Azure Application Gateway). Enterprise-supported offerings such as Red Hat OpenShift are now available in a managed services model across cloud providers.

These solutions address the complexity of administering and managing Kubernetes-based solutions, making adoption easier for enterprises. IT teams can focus completely on building business applications and less on cluster management, software upgrades, certificate management, and other complexities around Kubernetes.

In collaboration with Infosys, a U.S.-based specialty auto parts provider built Next-Generation Auto Repair Assistant Platform. It was developed using container technologies on AWS with ECS Fargate, along with Amazon's Simple Storage Service (S3), API Gateway, and DynamoDB. The platform helps car owners book services at registered garage workshops with the integration of Google Maps and geolocation. It leverages an AWS ECS Fargate cluster for its API back end, which completely removes any provisioning requirement and scales to workload, optimizing hosting costs.

Platform as a service

Trend 2: Serverless solutions become mainstream with evolving offerings

Function as a service (FaaS) is a serverless way to execute modular pieces of code on the edge of computer systems. FaaS lets developers write and update a piece of code on the fly that can be executed in response to an event, such as a user clicking on an element in a web application. Offerings such as AWS Lambda, Azure Functions, and Google Cloud Functions support several programming languages and even container images. Newer serverless offerings have emerged, including containers, messaging middleware, big data processing, workflow orchestration, and low-code (LC) platforms. In fact, serverless and LC together significantly increase the speed of application delivery. From the developer's perspective, there is no server and no need for extensive programming to create business logic. The developer tooling, software development kits (SDKs) and support for DevOps, observability, and debugging have also improved. Hyperscaler solutions continue to address challenges around provisioning scale, integration with other cloud services, and security.

Enterprises should view the space as an opportunity to streamline API, microservice, and event-driven application implementation. Serverless solutions will continue to drive down application costs and evolve to better suit varying workload requirements.

A leading engine manufacturer in North America partnered with Infosys to modernize its legacy approval system with a centralized, one-stop, and on-the-go approval solution. It utilized serverless technologies, including API Gateway, Lambda, Step Function, SES, SNS, Aurora serverless, S3, and Azure AD, to track and act upon asset requests from multiple heterogeneous systems in near real time. This resulted in significant licensing costs savings on the legacy database and forms software. It helped build a pay-per-use model for an application with a highly variable load.