Business Applications powered by AI are revolutionizing the customer experience, accelerating business speed, and driving employee productivity. According to Frost & Sullivan’s 2024 Global State of AI Report, 89% of companies believe AI and machine-learning will help them increase revenue, boost efficiency, and improve the customer experience.
Take Vodafone as an example. The telecommunications provider is using Azure AI services such as Azure OpenAI to deliver hyper-personalized, real-time experiences across all its customer touchpoints. This includes its digital chatbot TOBi. Naga Surendran is senior director of marketing for Azure Application Services, Microsoft. She says that Vodafone has been able to resolve 70% first-stage inquiries using AI-powered digital channels. It has also increased the productivity of its support agents by giving them access to AI capabilities similar to those of Microsoft Copilot – an AI-powered productivity software.
The result is a 20 point increase in net promoter score, he says. “These benefits are driving AI infusion into all business processes and applications.”
But realizing measurable value from AI powered applications requires a different game plan. Legacy application architectures are simply not able to meet the high demands of AI enhanced applications. To remain competitive, organizations must modernize their infrastructures, processes, and applications architectures with cloud native technologies.
It’s time to modernize
Organizations today are living in a world of geopolitical changes, increasing competition, supply-chain disruptions, and changing consumer preferences. AI applications can support innovation, but they must be able to scale as needed. By modernizing applications, organizations are able to achieve the agile development and scalability needed to support rapid innovations and accelerate the delivery AI applications. David Harmon, director for software development at AMD, says that companies “really want” to be able to migrate their existing [environment] to take advantage of the hardware changes.
Modernizing applications, data and infrastructure can improve customer experience. Consider Coles, a supermarket in Australia that has invested in modernization, and is using data, AI, and other technologies to deliver dynamic ecommerce experiences to their customers, both online and in store. Coles has reduced build times by several hours and switched from monthly to weekly application deployments with Azure DevOps. Coles was able to provide a more personalized experience for customers by aggregating customer views across multiple channels. According to a report by 2024 CMSWire Insights, AI is being used more and more in digital customer experience toolsets. 55% of organizations are already using AI to some extent, while others are just beginning their journey.
However, even the most carefully-designed applications are vulnerable to cyber attacks. Bad actors can extract sensitive data from machine learning models, or maliciously inject AI systems with corrupted data if they are given the chance. Surendran says that AI applications are now interacting directly with your core organization data. “It is important to have the right guardrails to ensure that the data is secured and built on a system that allows you to do this.” The good news, modern cloud-based architectures, can deliver robust security and data governance, as well as AI guardrails such as content safety, to protect AI applications against security threats and to ensure compliance with industry standard.
The solution to AI innovation
A new approach is needed to modernize applications. New challenges such as demanding customers and malicious hackers require a new way of thinking. Surendran says that you need the right application architecture in order to keep up with the market, and to bring applications to market faster. “Not having this foundation can slow you.”
Enter the cloud native architecture. As organizations adopt AI to accelerate their innovation and remain competitive, it is becoming more urgent to rethink the way applications are built and deployed on the cloud. By adopting open source software and Linux, organizations can create a flexible platform that is optimized for cloud and AI. Harmon explains open source software gives you options. “And the open source ecosystem thrives off of that.” It allows for new technologies to be brought into play.
Application Modernization also ensures that AI applications have optimal performance, scale and security. Modernization is more than just moving workloads from on-premises virtual machines to the cloud. Cloud native architectures are designed to give developers the following features.
- The ability to scale to meet changing needs
- A better access to data to power intelligent apps
- Easy access to tools and services for building and deploying intelligent applications
- Security embedded in an application to protect sensitive information
These cloud capabilities help organizations get the most value from their AI applications. Harmon says that at the end of day, it’s all about performance and security. Cloud is no different.
Surendran also notes that “when organizations leverage a cloud-based platform for modernization they can gain access to AI model faster and get to the market faster with AI-powered apps.” The modernization journey is driven by these factors.
Best practice in action
To reap the benefits of modernization, organizations must take steps to ensure both technical and operational success. These are:
Prepare employees for speed. As the modern infrastructure accelerates AI-powered application development and deployment, developers must prepare to work faster and more intelligently than ever. Surendran warns that employees must be skilled in the modern application development practices in order to support digital business needs. This includes developing expertise with loosely coupled Microservices in order to build scalable, flexible applications and AI integration. Start with an evaluation. Surendran says that large enterprises will likely have “hundreds, if not even thousands” of applications. In order to modernize their applications, organizations need to take the time and evaluate their application landscape. Surendran continues, “Starting with a thorough assessment is crucial.” “Understanding and taking inventory of the various applications, which teams are using what, as well as what this application drives from a business processes perspective, is critical.”
Concentrate on quick wins. The modernization of applications is a long-term, massive transformation in the way companies build, deliver and support them. Most businesses are still developing and learning the right strategy for supporting innovation. Surendran suggests focusing on quick victories while also working towards a larger application estate. “You must show a return-on-investment for your organization and its business leaders,” says Surendran. You can, for example, quickly modernize some apps with replatforming and then add AI capabilities.
Partner up. Surendran says that modernization can be intimidating. The first step is to select the right strategy, platform, and process that will support innovation. Organizations must also “bring in the right partners to help them through change management and execution of this complicated project.”
Addressing all layers of security is important. When it comes to protecting data, organizations must not be complacent. Surendran says that this requires a multi-layered approach to security, including: security by design (where products and services are designed with security in mind); security by default (where protections are present at every layer, interaction, and location where data is located); and security by ongoing operation, which involves using the right dashboards and tools to govern applications through their entire lifecycle.
Looking to the future
The majority of organizations are aware of the importance of application modernization. With the arrival of AI, modernization efforts will need to be done correctly and AI applications must built and deployed in order to have a greater impact on business. Adopting a Cloud Native Architecture can help by providing a platform that enhances performance, scalability and security, as well as ongoing innovation. Surendran says that as soon as you upgrade your infrastructure to a cloud platform you can access these rapid innovations in AI. It’s all about being able continuously innovate with AI.
Read about how to accelerate app- and data-estate readiness for AI innovation using Microsoft Azure and AMD. ExploreLinux on Azure.This content was created by Insights – the custom content arm at MIT Technology Review. This content was not written by MIT Technology Review.