Enterprise technology has always evolved in waves - from on-premise systems to the rise of SaaS, and now, into the agentic era of AI. Each shift has reshaped not only how organisations run but also who controls the levers of transformation. According to Chaitanya Peddi, Co-founder, Darwinbox - HR is at the very center of this transformation. With AI agents becoming a universal interface across enterprise systems, HR holds the decisive advantage: it manages the most critical context, governs access controls, and is the only function that touches every employee. In this shift, HR is not merely a participant in technology adoption; it is the architect of how enterprises will function, scale, and thrive. But unlike earlier shifts, this transition is not just about software - it is about redefining the very architecture of work. 

From Fortresses to Flexibility: The SaaS Breakthrough

Before the 1990s, enterprise technology meant manual processes tied to rigid, on-premise systems. Organisations had to custom-build software, invest heavily in servers, and maintain large IT teams. 

“Imagine an organisation as a walled fortress. Adding a room isn’t as simple as placing bricks; it often requires dismantling existing structures and rebuilding from within. Similarly, implementing a new policy or designing a fresh workflow typically demands intervention from IT. The process is rarely swift; it unfolds on their timeline, with layers of complexity and dependencies,” Chaitanya explained in a recent keynote streamlining the history of enterprise technology. 

This was the rigidity that gave rise to “Software as a Service” or SaaS. Everything moved to the cloud.  “So when things move to cloud, you don't need massive infrastructure investment, upfront investment, so all you need is to see the software that suits your requirements, buy that off the shelf, and implement and go live,” added Chaitanya. Business functions could now purchase cloud-based tools “off the shelf” and go live with minimal infrastructure.  

With the rise of SaaS, functional teams gained greater autonomy over their tools and systems, a welcome shift from centralised control. However, most enterprise software was still dominated by large vendors and legacy ERPs. When the cloud revolution emerged, these incumbents were slow to adapt, taking years to migrate their platforms fully. This inertia created a window of opportunity for agile innovators to build targeted point solutions. By focusing on specific pain points, they developed lightweight SaaS tools tailored to individual functions. As a result, today, nearly every business function operates its own specialised software ecosystem.

This introduced a new challenge: proliferation. Each department, and sometimes even sub-functions, adopted its own solutions. This created a sprawl of disconnected platforms - what solved rigidity, now created fragmentation.

From Flexibility to Fragmentation: Logic Debt

This proliferation produced what many call logic debt. Each tool operates on its own underlying logic, which means seamless integration is far from guaranteed. Consider systems like HRMS, payroll, and CRM: though they all interact with the “employee object,” their definitions and data structures often differ significantly. What one system identifies as an employee may not correspond with another’s interpretation. As organisations adopt more tools, these inconsistencies multiply, making integration increasingly complex and cumbersome. “The very issue SaaS was meant to solve ended up being amplified instead,” Chaitanya pointed out.

Entire industries emerged just to connect these systems. Integration platforms like MuleSoft and Workato became indispensable, but the underlying issue was that a fragmented enterprise experience persisted.

It is not unusual to see HR teams running on 30 tools these days - from finance, CRM, and supply chain functions to hundreds more. Integration itself became an industry, but the fragmentation felt unsolvable. 

With the advent of Artificial Intelligence, this problem finally found a solution. So, how is AI solving the problem? 

From Fragmentation to Flow: The Role of Reasoning AI Agents

According to Chaitanya, the first phase of AI, which was content generation and summarisation, is already over. The next phase is reasoning: enabling systems to understand intent and dynamically complete actions. This is where AI agents play a major role in solving the issue. 

At its core, any software allows users to interact with a database - retrieving or writing data. Traditionally, this required navigating a UI layer and following a fixed workflow.

Agents changed that model. An agent is something that has access to some information or a tool. A tool either fetches the data or completes one action. Earlier, this reasoning layer was left to the users. Now the agent does it dynamically. 

“So, an agent is essentially a system that has access to a set of tools. You are asking the agent to complete this action, and you are giving the agency. So the best part is, for an agent, you take any system, whether it is HR or CRM or a finance or ITSM, no matter how complex, it sees any system as just a bunch of the tools, for example, a HCM, whether it is Darwinbox or any other system for a agent, it is probably bunch of, you know, 100, 200 tools. So for an agent, any system is just an aggregation of tools,” explained Chaitanya. 

Earlier, this reasoning layer was left to the users. With the language models, the reasoning is something that got added. “Provided it has access to all the tools from your finance system, CRM system, project management system, and HR system, the  AI Agent will reason out that query in what sequence those two tools need to be executed, and complete the task,” Chaitanya said, highlighting instances of this at work at Darwinbox.

An employee asked an HR bot to recommend vacation dates that would minimise leave usage. The agent accessed the holiday calendar, leave balances, and teammates’ schedules, then applied for leave on the employee’s behalf.

“So, before agentic, if we wanted to solve this use case, we’d have to train the system to solve it. Step one: get this data. Step two: get this data. Step three: get this data. Then you write an algorithm, OK, this is how you optimize. Then you recommend. So now there is nothing, so you just, you know, give access to the tools and ask, you know, what you want, it reasons out. So the brain part got added to the entire set of capabilities,” explained Chaitanya.

In another case, a project manager used the same bot to interact with JIRA — retrieving tickets, reassigning tasks, and updating statuses. A salesperson could query their HubSpot pipeline in the same way, all without leaving the HR interface.

Each system becomes a set of tools the agent can reason with - eliminating the need to switch between applications.

Why HR Could Lead the Agentic Era

The question is no longer if broad enterprise agents will exist, but which departments will be spearheading it. “So whether we agree or accept or not, this is the direction enterprise tech is heading towards, whether six months, one year or two years, only time will tell, but you will get to a stage where all the companies will build a generic, broad agent, and most the population will be interacting with different systems through that agent, not through UI anymore,” observed Chaitanya. 

So now what is interesting is who will become that generic agent - ITSM is pushing through CIOs, sales force is pushing through CMOs and CROs, Microsoft and JIRA, for instance, are pushing through CTOs - to become that agent. But Chaitanya feels, here HR has some distinct advantages - “There is a natural advantage for HR & CHROs to win in this game. Because HR has ownership of the most important context,” Chaitanya shared. HR system has the potential to become a broad agent in this regard.

“There is one system that touches everyone today. So if you want to build a broad agent for everyone to access, the HR system has a natural advantage. It is a simple natural extension,” shared Chaitanya. 

  • Access control: Most important access controls are defined based on the context the HR system manages — your location, grade, department, etc.

  • Universality: HR systems touch every employee, unlike CRM or finance platforms used only by subsets of staff.

  • Trust: Employees trust HR systems more than finance or CRM to handle personal data. “Studies clearly indicate people generally trust the HR system, and HR a lot more from access control, for example, if you want to give access to your personal calendar, apparently, the study only says it's not what I am saying. So people trust the HR system a lot more than a finance or a CRM system to actually give access to their personal stuff as well. So I think you know that is the direction. So it is where we are heading to, and there is a clear precedent,” Chaitanya said. 

This makes HR not just a user of agents but the natural owner of the enterprise’s central agent. 

Some organisations are already signalling this shift. At Moderna, the CHRO also serves as Chief Transformation Officer, with the CIO reporting into that role. Similar models are emerging at Novartis and Schneider.

Workforce planning is increasingly reframed as work planning. “I feel this is the start of a very interesting era, because, at the end of the day, workforce planning will simply become work planning. There is work, and someone has to figure it out. So who is better equipped to do that work? Humans or agents. And I feel HR is going to be the guardian of that thinking. Hence, you need to lead that transformation,” Chaitanya speaks, urging CHROs. 

As enterprises move toward broad, generic agents through which most employees interact with systems, for HR leaders, this is both a responsibility and an opportunity. By leveraging their universal reach, contextual access, and trust advantage, they can position themselves as the stewards of the agentic era. 

Darwinbox’s ambition is to support HRs to lead the transformation. “So that is the direction the enterprise tech is going towards. So definitely, whether it's one year or two years, it will be a broad agent, and there is a clear precedence that HRs are going to manage the transformation and become Chief Transformation Officers, along with the CHRO role, Chief People Officers.” 

From fortress-like on-premise systems to SaaS, every technological wave has redistributed control within enterprises. The agentic AI era will be no different. This time, however, HR is uniquely positioned to sit at the centre, not only shaping how employees interact with technology, but also driving organisational transformation itself.