Service
Artificial Intelligence
Building a Smarter, Connected Tomorrow with AI-first and Agentic AI
Cognitive Business Operations
Unlock Business Agility and Future-ready Operations with TCS
Highlights
- The document discusses the transformative role of Artificial Intelligence (AI) in application maintenance, emphasizing its potential to enhance system reliability, service delivery efficiency, and customer experience.
- It highlights five key trends, including intelligent business observability, AI-driven troubleshooting agents, and the use of conversational agents and copilots to improve user experience and engineering productivity.
- The document also explores the concept of an intelligent enterprise knowledge fabric and the future of autonomous IT operations powered by AI.
On this page
On this pageinpage
The era of AI dominance in Application Maintenance
In an era characterized by an unyielding demand for ‘never-to-fail’ IT systems and diminishing budgets for application maintenance, the growing influence of Artificial Intelligence (AI) in this space is proving to be an inflection point. TCS believes that embracing AI in application maintenance can transform the reliability of systems, enhance the efficiency of service delivery, and reimagine the experience of service consumers. TCS believes that AI can play a role in building an intelligent knowledge fabric for an enterprise, that forms the foundation for AI algorithms to run effectively. Production data managed by application maintenance teams contain essential insights into business processes and customer journeys and can be used to improve operational efficiency. This paper highlights five trends on how AI could play a defining role in shaping the future of application maintenance.
Intelligent business observability
Application Maintenance has historically been about managing health of IT applications. Production data holds valuable information, not just about health of IT systems, but also on experience of end-user through their digital journeys, real-time progress of business process flows, and much more. TCS is working on combining this data with AI, to help organizations transform their businesses for efficiency and experience, the power of which is illustrated below with examples.
Ensure seamless customer journeys: Business observability, accomplished by hooks to production data across the process flow, can ensure processes are completed successfully and on time. A manager at a manufacturing company would be able to get great control over their Order-to-Cash flow; a trader at an investment firm would be able to trace a transaction or order, correlate events across trade desks, risk engines and clearing houses to detect latencies and errors across processing lifecycles, and ensure issues are detected and averted.
Taking it a step further, business observability will help assist customers in their journeys in real time. Imagine a bank that calls up its customer stuck at shopping mall with a failed card transaction at a POS terminal and guiding them through to successfully complete the purchase, or a web shopper being assisted by an online agent, if found to be struggling to make choices.
Enhance business efficiency: Business observability will help in analyzing business performance, thereby providing ways to improve efficiency. For example, real time data from stores and warehouses can help a retail company improve its supply chain efficiency. Insights into customer preferences and usage patterns can guide smart merchandising and personalized marketing campaigns. Pricing strategies can be derived from insights from transaction volumes.
Insights to strengthen business operations: Business observability can bring insights into events that are developing in the ecosystem and make our systems more secure and reliable. For example, detecting anomalous patterns in user login and errors in access logs can highlight a potential security threat. Market events can be predicted from sudden spike in system throughput and latencies caused by unexpected transaction volumes. Observability can correlate failed regulatory reporting jobs with data lineage and workflows and alleviate business risks.
Balancing autonomy and reliability with swarm intelligence
Autonomous IT Operations is spoken of as the future of application maintenance. TCS’ Agentic AI solutions for application maintenance leverages the concept of swarm intelligence to enable a set of agents that perform tasks, communicate and coordinate with each other, evaluate the health of systems post automated actions, learn from real-time developments, and course correct as needed to reach the overall objective.
In the context of incident management in application maintenance, imagine a scenario where there are agents to accept tickets, auto-route it to the best engineer to act, collect logs and knowledge articles, and create insights and intelligence to diagnose, and further recommend a set of steps for resolution. For proven scenarios (that are learnt from observation by AI), the actions can be autonomous. However, for scenarios where the confidence of resolution is less than expected, humans can make the final decisions before actions are performed. Agentic AI, with orchestrated flow of agents, but balanced by human controls, will deliver substantial efficiency, while ensuring reliability.
Navigating the complexities of investigating application issues with AI
Imagine a scenario where a customer disputes the tax or rebate computed on an invoice, this would require the maintenance engineer to understand the application logic, data processing flows, and computations involved. This requires a good understanding of functional and contextual knowledge, combined with an appreciation of technology. This makes it time and effort consuming to analyze and resolve. TCS envisions that the use of AI and Gen AI technologies will significantly ease the life of the service engineers to analyze different elements of IT landscape.
TCS’ work on leveraging AI to transform service engineer productivity in handling investigative tickets include:
Talk-to-logs solutions: These help in analyzing logs at every layer - application, infrastructure, and database by leveraging AI to extract insights and intelligence to analyze the root-cause and identify potential solutions.
Talk-to-code solutions are complex but will be the need of the hour to handle complex investigative issues involving business logic. These agents using AI will assist in extraction of business logic from code for analysis, analyze the impact of code changes on the overall functionality, identify technical debt, and potential vulnerabilities.
Talk-to-data solutions help trace the flow of data from source to target and identify the step in processing flow that needs to be analyzed to resolve the issue. Capabilities to extract data from data sources using natural language (ex- Text to SQL) could significantly assist the debugging process.
Integrated intelligent resolution recommenders, eventually, combine the intelligence derived from the analysis of logs, code and data to provide step-by-step guidance to engineers to take next steps shall significantly drive down the ticket resolution time.
Transform user experience with conversational agents and copilots
It would be no longer required for users to raise tickets and wait for a response, after its assigned and acted upon. TCS’ Gen AI powered solutions will give the end user the ability to converse with agents and get their requests fulfilled.
Conversational agents are expected to influence a range of stakeholders within the Application Maintenance ecosystem, including end users, business users, technology managers, and service engineers.
The end user might need assistance to create a ticket or check status of an existing one. The user might need quick attention to gain access to a resource or to reset a password or seek clarifications for a query on a report or service that he/she has received. A virtual conversational agent can transform the experience of the end user by providing contextual responses to them in these scenarios in quick time.
A business or technology leader might want to review the performance of their business processes or IT systems at their convenience. As elicited earlier, business process performance, and end user experience are critical success factors and governance of these are of paramount importance. A virtual agent can provide instant access to reports, insights from analysis, and could customize them as needed.
Likewise, service engineers need help to analyze issues and take necessary steps to resolve. Digital Assistants and Co-Pilots shall make the life of service engineers easy and productive.
Intelligent Enterprise Knowledge Fabric: The foundation to reducing human dependency and enabling autonomy
Moving towards autonomous, AI led operations would need a strong foundation of codified knowledge that intelligent systems can rely on. Historically, there have been experts in the ecosystem who have gained knowledge through years of service experience and scattering knowledge documents like process manuals, system documentation, runbooks and standard operating procedures. Reliance on experts can result in dependencies and risks, and the presence of unstructured and disparate knowledge sources may be tedious for engineers to consume. TCS’ Intelligent Knowledge Fabric solutions, powered by AI, is a game changer in knowledge management. A few of its key features include:
Intelligent Search: AI combined with graph technologies helps in connecting disparate knowledge sources, existing in different parts of the enterprise and in different file formats in creating knowledge graphs that integrate the content into a single repository. Knowledge graphs can be searched using Gen AI through easy prompts to get contextual responses to queries communicated in natural language. This would have otherwise required manual search through several documents.
Generate Knowledge, not just consume: AI has the power to review the availability of knowledge and identify areas where we need to enrich. AI can be used to identify scenarios of incidents and failures from ticket logs and then use historical information to generate knowledge for each such scenario.
Extract knowledge from experts: AI can be used to transcript conversations, discussions and working sessions with experts and leaders. AI can then process these transcripts to generate knowledge assets and integrate into the overall knowledge fabric.
Learn from production behavior and embellish: Application behaviors change over time and it’s important to re-baseline performance and learn to adjust thresholds and capacity requirements. Knowledge agents are expected to pick up these clues and re-configure the functioning of other AI agents in the ecosystem.
Looking ahead
The road ahead is exciting with endless possibilities, with the emergence of AI in mainstream adoption. TCS has evolved a comprehensive and contextual framework – ‘Application Reliability Engineering’ which helps organizations to reimagine application maintenance with the power of AI. The framework has evolved with over 150 assets, comprising of AI agents, agentic workflows, benchmarking toolkits, and process handbooks to guide the execution teams through their transformation journey.