From Manual Help Desks to Digital Ticketing
Until recently, IT help desks operated entirely through manual processes. When an employee or customer encountered an issue, the standard course of action was placing a phone call or sending an email, followed by a prolonged wait as a support agent manually recorded the problem and began working toward a resolution
As organisations first adopted PCs in the 1980s and 90s, help desks were set up to respond to users’ technical problems. Long resolution times were common, and support was strictly human-only. Agents often juggled repetitive tasks like resetting passwords or routing tickets, which bogged them down and led to backlogs. In those days, being “on hold” with IT support was a shared frustration (1).
Things began to change in the late 1990s and early 2000s with the rise of digital ticketing systems. Email and web portals started to supplement (and sometimes replace) phone calls for logging issues. IT support organisations adopted frameworks like ITIL, and “help desks” evolved into “service desks” (2) – a term change meant to signal a more organised approach to IT support. Instead of scribbling issues on notepads, companies introduced centralised ticket management software (think early BMC Remedy). These systems allowed for better tracking of issues and basic automation, like assigning tickets to the right team or sending notification emails. This evolution improved organisation, but the actual support workflow was still very much human-driven.
The First Wave of Automation
As service desks became digital, organisations looked for ways to handle growing ticket volumes more efficiently. Around the 2000s, teams started building out knowledge bases so that answers to common problems could be documented and reused. Simple scripts and workflow rules were introduced to handle routine steps. For example, automatically routing tickets based on keywords, or sending an automatic reply acknowledging a request. These early automations were relatively basic, but they laid important groundwork (3).
Email had become a dominant support channel by the 1990s, reducing reliance on phones. By the late 2000s, many service desks offered self-service portals where users could search FAQs through a catalogue. Password self-reset tools (or SSPR) became common, offloading one of the most frequent IT help requests. Service desk software added capabilities like workflow automation (to enforce processes and approvals) and integrated remote support tools to fix machines without leaving one’s desk. All of this helped cut down resolution times and eased some strain on support staff, but the service desk was still largely reactive, which meant waiting for someone to report a problem.
AI Enters the Scene
Fast forward to the 2010s, and service desks began embracing a new helper: Artificial Intelligence. Early implementations were in the form of chatbots. These were basic virtual assistants that could greet users and answer very simple questions. Initially, these bots were rule-based, handling FAQs like “How do I reset my password?” with scripted answers. They were useful but limited. Over time, though, they evolved to more sophisticated AI-powered assistants that could understand phrased questions and pull answers from a knowledge base.
Consequently, AI started acting as a copilot for support technicians. Research in this period introduced machine-learning tools that could learn from historical ticket data to assist with routine support tasks (4). Naturally, these tools reduced the burden on human agents, who were tasked with repetitive, often boring work.
By the late 2010s, forward-thinking IT teams had integrated AI chatbots into their service desks to handle a sizeable chunk of Level 1 support. In essence, AI became the new front-line support rep, working alongside human agents as a tireless junior colleague.
The Rise of Agentic AI
Today, we are witnessing a new era: one powered by autonomous, agentic AI capable of operating with minimal human intervention. In essence, these AI systems are moving beyond reactive resistance and starting to act independently to resolve issues. Unlike a traditional chatbot that only provides information, agentic AI has the potential to execute multi-step tasks in response to inputs from users in natural language.
Think of the difference in scope like this: A user asks, “My laptop is running really slow lately”. A traditional bot might respond, “I’ve created a ticket for IT to look into your issue.” An agentic AI, on the other hand, could say, “I see your laptop is low on storage. I’ve taken the liberty of clearing the temp files and ordering you an external drive and scheduled a backup.” This kind of AI has the autonomy and integration to not just talk but also to take action on the user’s behalf.
Of course, this kind of response is still aspirational. In most enterprises, AI systems do not yet have the authority to access end-user devices or carry out procurement-related tasks independently. For agentic AI to operate at this level, robust integrations with endpoint management platforms, procurement systems, and well-defined governance frameworks are required. . While this level of functionality is not yet widespread, the foundational technologies are maturing rapidly, indicating a clear trajectory in that direction.
The difference is that agentic AI, built on large language models, can reason, summarise, write, and analyse unstructured data. It can use inputs like documents, excel sheets, or chat logs to understand the context of a problem and tailor responses on a case-by-case basis. The potential is huge, and the key advancement is that these AI agents can improve over time by learning from feedback and outcomes.
This is a big leap from earlier automation, which only did exactly what it was programmed to do. With agentic AI, the service desk starts to resemble a living system; one that adapts and gets smarter with each ticket and conversation.
In this context, service desks are also beginning to explore proactive and predictive capabilities such as anomaly detection and pre-emptive ticketing. For example, some modern platforms can flag abnormal system behaviour (like a spike in memory usage) and automatically create a ticket or even trigger an automated response before a user reports the issue. While not yet mainstream, these capabilities are actively being explored and piloted by enterprises.
A New Era for Service Desks and CTOs
We have gone from filing paper trouble tickets and waiting ages for a callback, to having AI assistants at our beck and call, resolving issues in seconds. This evolution hasn’t happened overnight, but the acceleration in recent years has been astounding (and it is only just the beginning).
Embracing AI in service desk is not just about cutting costs and saving time, it is about delivering a better experience for employees and customers. It means IT support can scale effortlessly as the company grows, and it can adapt quickly to new issues by learning from data.
Some organisations have started seeing early outcomes from this transformation, such as shorter resolution times and higher user satisfaction. Predictive systems that generate tickets automatically for recurring issues have been shown to reduce support workload and improve uptime.
There are considerations to get it right, of course. You need quality data for AI to learn from, and ongoing training and tuning of these systems. Change management is key. And maintaining a human touch for complex or sensitive issues remains vital. AI augments humans, it does not replace the need for empathy and creative problem-solving.
For tech-savvy leaders, it is an exciting time to reimagine what IT support can do. The evolution is underway, and the service desk of tomorrow is already starting to take shape today.
About the Author:
Himavanth Dore is a Product Manager at Wolken Software, where he leads the strategic planning and execution of cross-functional initiatives that drive enterprise innovation. With a focus on delivering high-quality outcomes aligned with organizational objectives, he ensures seamless collaboration across teams, optimizes resource utilization, and proactively mitigates project risks. Himavanth serves as a key liaison for stakeholders, promoting transparency and consistent communication throughout the project lifecycle. His expertise in integrating technical and operational components, coupled with agile responsiveness to evolving requirements, enables him to deliver impactful and scalable product solutions.
About Wolken Software
Wolken Software is an award-winning enterprise-grade B2B SaaS company that uses agentic AI to automate enterprises’ workflows. With its 3S Model (Simple, Scalable, and Secure), Wolken Software empowers global clients in their digital transformation journey. It offers Enterprise Management and Customer Service Desk by providing a powerful, low-code platform to streamline business processes.
Wolken Software caters to customers from the banking and financial services, semiconductor, software, consumer goods, and electronic component industries and has grown to add many Fortune 100 companies to its clientele in the US, Europe, and Asian markets. It supports over 7,000 agents serving more than 50 million active end-users across 60 countries and processes more than seven million tickets annually.
1) For example, at Glaxo Pharmaceuticals the help desk was originally headed by one senior analyst using personal expertise to solve user problems. There was little automation or knowledge tooling, leading to repetitive inquiries and slow resolutions. In fact, in 1988 when Glaxo introduced an expert system “Rupert” to assist the help desk, it cut the average time to resolve user problems by half.
2) While the terms “help desk” and “service desk” are often used interchangeably, most agree that the help desk originally focused on technical issue resolution, and gradually evolved, particularly with the rise of ITIL, into the broader and structured service desk model of today.
3) Initially, knowledge bases were simply a self-service library of articles, guides, and FAQs to assist users and support agents in resolving common issues. Over time, they have transformed into sophisticated knowledge management systems, with various methods and tools to manage organisational knowledge effectively.
4) Scholars in the field developed AI models to auto-classify incoming tickets into categories or priority levels and to suggest relevant knowledge-base articles or past solutions that match the user’s issue. Some AI systems could even draft an initial response to the user by pulling information from similar resolved cases.