AI-powered Application Evolution Services

Application Managed Services can help companies outsource the operation and continual development of applications, such as ERP systems, to service providers. The range of services typically extends from troubleshooting, updates, and performance improvements to end-user support and ongoing optimisation of applications throughout their entire lifecycle.

Common AMS practices are largely determined by Information Technology Infrastructure Library (ITIL) principles. They promote the consistency, efficiency, and quality of AMS by creating standardised processes and clear responsibilities:

• Service design enables AMS to meet business requirements and is scalable, secure, and cost-efficient.
• Service transition structures the shift to new or changed services. Additionally, it enables change management, reduces risks, and speeds up implementation.
• Service operation enables stable operation of applications through incident, problem, and event management, as well as effective support processes.

Alignment with such standards helps deploy IT capacities in a targeted and cost-efficient manner, reduce errors and security risks, and transparently monitor service quality through clear KPIs and SLAs.

Challenges with traditional AMS approaches

Despite the structured approach based on established standards and leading practices, traditional AMS presents many challenges. For example, support teams usually are not sufficiently familiar with business processes, making it virtually impossible to develop application operations in a business-centric manner. In addition, automation opportunities are not fully exploited. Many AMS providers employ resource-based models that are handled manually, leading to inefficiencies and high operating costs. Purely reactive support via ticketing on a widespread basis results in delayed and lengthy resolution times and a poor user experience.

Ongoing digital transformation offers numerous opportunities to overcome the weaknesses of traditional AMS. Recent significant advancements in agentic and AI-based solutions have created the ability for new AMS/AES providers to embed AI into their delivery model, disrupting the traditional AMS providers in the market. Advanced cloud technologies increase scalability and security. Methods such as predictive analytics help to identify problems proactively before they occur.

The idea behind Application Evolution Services

Application Evolution Services (AES) is an approach developed by PwC that aims to further develop AMS in a business-centric manner based on the opportunities offered by a combination of human and digital workers. AES champions AI-enabled operations, optimisation, and innovation that go beyond reactive maintenance and support. We’ll continuously improve applications, monitor their infrastructural and organisational framework conditions, and actively adapt them to the needs of the business, by running operations faster and smarter.

Elevating AES with AI agents

With Agentic AI, a new type of artificial intelligence is gradually maturing that is virtually predestined for use in AES. Key features of Agentic AI include a higher degree of autonomy and the ability to reliably perform even complex tasks. This is implemented by breaking tasks down into sub-problems, each of which is solved by suitable AI models. In addition to individual AI agents for specific tasks, multi-agent systems are therefore becoming increasingly popular, in which a higher-level AI agent orchestrates several subordinate AI agents.

The first digital agents are already in use at AES. They automate workflows, generate suitable user stories in DevOps scenarios based on requirements, conduct change impact assessments, and relieve the traditional help desk by answering user queries via chat. The potential applications are vast. AI agents are helping shift AMS from a reactive support model to a proactive, continuous learning service that dynamically adapts to challenges and business goals. 

• AMS are based on ITIL principles and leading practices and promise clearly defined service quality through KPIs and SLAs.
• Traditional AMS approaches are highly human based with limited automation. They are usually limited to purely reactive ticketing support delivered by lower skilled resources. They do not address business processes and are still characterised by manual procedures in many areas.
• AES is an industry-infused, business-centric approach that proactively improves applications and develops them further based on business requirements. In addition to maintenance and support, it focuses on AI-enabled operations, optimisation, and innovation.
• AI agents open up new possibilities for AES to improve workflows, reduce manual effort, and improve service quality. They are already in use today—for example, in areas such as helpdesk and workflow orchestration.

AI is reshaping AMS in diverse ways. Below, we explore the key areas where AI is making a difference, showcasing real-world examples of AI agents that currently enhance AMS operations.

IT Operations Management (ITOM)

AI agents offer exciting potential for IT operations management by automating Service Desk services, auto resolving incidents and events more quickly, and identifying recurring problems. The prerequisite for using agents is the integration of metadata, process data, and runtime information—for example, server and application logs, information on software versions, maintenance logs. 

Based on a thorough, consolidated database, agents can take on specific monitoring tasks and also control existing AI-supported processes. These can include:

• Process intelligence tools that help companies analyse business processes and improve workflows
• Methods such as predictive analytics and anomaly detection, which can help identify optimisation opportunities at an early stage and avoid service interruptions 
• AI-driven warning systems that enable faster and more accurate responses to incidents based on real-time monitoring.

Support and customer service

The use of chatbots and virtual assistants helps provide seamless end-user support while reducing the workload of human support staff. These self-service systems are available around the clock to answer frequently asked questions and solve simple problems. Voicebots can also interact with users by phone, identify their concerns, and provide immediate assistance or connect them to the right contact person.

Based on the most frequently asked questions, AI systems can generate help content, such as FAQs that reflect current support needs, and thus speed up problem resolution. AI agents can suggest tried-and-tested solutions and, depending on the area and the configured permissions, even implement simple fixes themselves.

AI agents are helping redefine how application services are run, particularly by personalising customer service. By analysing customer data and past support requests, these agents provide tailored offers and solutions that directly address specific issues. Well-designed data management is essential for successful implementation. 

Workflow management

IT service management offers many opportunities to automate business processes with AI agents. They help to control and improve clearly defined workflows. Here are some areas that benefit from this:

• Automated ticket processing 
  AI agents automatically prioritise requests and tickets as part of intelligent incident management and forward them to the responsible teams. Depending on the setup, low-complexity issues can also be handled automatically to reduce the workload of support staff.
• Intelligent error diagnosis 
  AI agents recognise patterns in system data and proactively suggest solutions. For example, analysing server data can reveal anomalies in CPU or memory usage. The AI agent analyses the problem, alerts the operations team, and helps prevent operational disruptions.
• Compliance with SLAs 
  Continuous monitoring of processes enables critical incidents to be identified and prioritised more quickly to meet the response times defined in service level agreements. Predictive analytics also enables AI to suggest preventive maintenance measures, maximising availability and avoiding SLA violations.
  

The standardisation potential of digital agents

• Bundled knowledge management 
  Digital agents can consolidate data and insights from ITOM, customer service, process execution, including the potential for industry-specific insights. They centralise knowledge management and ensure that all stakeholders receive up-to-date and consistent information.

• Unified service experience 
  AI agents ensure a consistent service experience across departments and functions through automated process control and unified workflow tools. They create seamless transitions between service processes and improve collaboration between teams.

• Real-time decision support 
  Digital agents can continuously synchronise real-time data between systems. This enables that process data, customer interactions, and operational insights flow directly into decision-making processes.

• Analytics and reporting 
  By consolidating data from different platforms, AI agents lay the foundation for advanced analytics and accurate reporting. This enables trends, anomalies, and potential risks to be identified early on and operational processes to be improved in a targeted manner.

• AI agents offer immense potential for ITOM by proactively identifying bottlenecks, security risks, and events and enabling early optimisation of the IT landscape.
• Support and customer service benefit from AI agents through enhanced 24/7 self-service offerings. Virtual assistants can answer questions, solve simple problems independently, and relieve human support staff.
• In workflow management, agents can prioritise tickets, diagnose errors, and increase efficiency in IT service managent.
• By identifying critical incidents early and suggesting preventive measures, AI agents help ensure that SLA targets are reliably met.

AI agents can be used in many ways in AMS. This makes a clear strategic approach all the more important. We outline the main steps below.

Initial integration and setup

To exploit the cross-functional potential of digital agents, they should not be deployed in isolated systems. It is more effective to deploy them on all AMS platforms in use and integrate them into the existing IT infrastructure and management, from the service desk to ITOM to the CRM system. 

A crucial step in this process is the creation of a common knowledge base. With the help of established data pipelines, we can consolidate metadata, customer service records, business process documentation, operating logs, and other relevant sources into a data lake that provides the knowledge foundation for digital agents.

Operational optimisation

Once set up, the next step is to develop application operations with digital agents. The following measures are particularly suitable for this purpose:

• Predictive maintenance 
  By analysing runtime data, digital agents can identify application problems at an early stage, take preventive measures, and thus minimise downtime.

• Automated incident management 
  AI agents should be used to handle simple, recurring service requests autonomously. Bots can cover a sizable proportion of first-level support. More complex problems can be escalated to human support team members as needed.

Personalised customer experience

While the integration of agents into first-level support through constant availability is already a first step toward a better customer experience, further opportunities for personalised self-service are opening up. With the use of digital agents in customer service portals, personalised recommendations and solutions can be provided based on historical service data and individual preferences.

The key to a higher degree of personalisation is the integration of feedback loops. By collecting customer feedback across all interactions, digital agents can continuously refine the service offering and accelerate responsiveness.

Continuous process improvement

Based on real-time process information and performance metrics, AI agents enable the dynamic adaptation of AMS workflows. For example, service delivery is improved through better resource allocation.

Digital agents also simplify the creation of thorough reports and analyses on service performance, customer satisfaction, and operational efficiency. They make the status quo transparent—and highlight trends and opportunities for improvement.

Future scaling and adaptation

The use of digital agents should not only consider the current situation. IT infrastructure is constantly evolving, and agent-driven areas must be prepared for increasing service demands. The resources used must be seamlessly scalable across regions and service lines.

Even when bots are already in play, it’s crucial to encourage continuous innovation. AI technology is evolving rapidly. The challenge is to integrate these advances into AMS processes and customer experiences.

Use of digital agents requires a cross-platform and cross-system integrative approach. A central prerequisite is the aggregation of relevant data on which the agents operate.

• To improve operations, AI agents can use methods such as predictive maintenance to identify and address problems at an early stage. As part of automated incident management, bots take over large parts of first-level support.

• Digital agents can improve customer service through personalised recommendations and offers. This requires the implementation of feedback loops. 

• Based on process information and performance indicators, AI agents can dynamically enable AES processes, like a Change Impact Assessment, and increase operational efficiency.

• To be prepared for the future and growing service requests, AI-related resources should be scalable. 

The use of AI agents already offers tremendous potential in the context of AES. To benefit from this, companies should consider the following points in particular:

Success factors for the use of AI in managed services

Digital agents offer decisive advantages for moving from traditional AMS to AI-enabled, continual application evolution. They refine processes, deliver data-driven insights, increase customer satisfaction, and foster proactive service management. However, their successful integration isn’t guaranteed. It requires careful planning, targeted training on agent-driven workflows, and continuous assessments to harness AI’s full potential—especially in a constantly evolving technological landscape.

AI agents are already being used productively in many areas. They support more efficient IT operations, support more effective and responsive processes, and deliver a whole new customer experience. By bridging gaps among silos and establishing uniform processes across departments and functions, they unlock new opportunities for companies to drive growth and differentiate themselves from competitors.

Looking ahead, AI‑driven application evolution services will increasingly underpin how applications are run and evolved. We see AES becoming a foundational capability for modern IT operations —to keep up with the fast-paced IT landscape, continuous innovation and proactive IT operations adaptation are crucial. Those who have embraced AI in their service strategies are already seeing the outcomes of running faster, scaling smarter and leading stronger. Get in touch to explore how PwC’s AI‑driven Application Evolution Services can help you transform how your applications are run, evolved and scaled.