Infosys TechCompass

API platforms as a service

Trend 10: API-driven automation reshapes infrastructure delivery

Organizations are increasingly building internal PaaS-style platforms that expose APIs for development, continuous integration and continuous delivery/deployment (CI/CD), and infrastructure orchestration. These internal marketplaces provide a simplified, often opinionated abstraction over cloud or third-party SaaS services, while giving the organization control and insights into the use of these services, which can be critical when dealing with security vulnerabilities and compliance aspects.

A leading conversational AI company serving Fortune 500 clients implemented platform engineering capabilities for customer support and e-commerce chatbots, with support from Infosys. The solution included a self-service developer portal, automated infrastructure provisioning using IaC, and integrated observability tools for real-time monitoring. Standardized CI/CD pipelines and Kubernetes-based container orchestration reduced deployment time by 60%. This approach enabled faster feature delivery and improved reliability and scalability across multiple workloads.

API platforms as a service

Trend 11: Edge clusters mature into cloud-native runtimes

Edge clusters are maturing into lightweight, cloud-native runtimes. Organizations run small Kubernetes flavors (k3s and the like) or purpose-built orchestrators onsite to host containerized microservices close to users and devices. This improves millisecond-level responsiveness while keeping central control planes. Functions at the edge-driven and event-driven microservices let teams push only the minimal processing to edge clusters, lowering cold-start costs with prewarmed runtimes and speeding per request latencies for APIs. Frameworks and platforms catering to this space are provided by both major cloud providers (AWS IoT Greengrass, Azure IoT Edge) and the open-source space (Kubeless, Apache OpenWhisk, FaaS, Knative).

Telecom operators are deploying multiaccess edge computing nodes to support ultra-low-latency APIs, allowing service providers to act as regional cloud and edge operators. As AI moves closer to users and devices, micro clusters equipped with graphics processing units and application-specific integrated circuits support use cases such as video analytics and predictive maintenance, running inference at the edge to maintain predictable response times when a high-bandwidth link to a cloud data center is not feasible.