When Satya Nadella declared that “SaaS is dead,” it sounded like provocation. But in the unfolding age of agentic automation, the idea might not be that far-fetched. Software, as enterprises have known it for two decades, is morphing into something fundamentally different. Something that learns, acts, and reasons in ways that break the boundaries of traditional apps.
In this transformation, it is not just technology that is changing, people too, are undergoing a shift. As automation systems evolve into intelligent agents, the nature of IT work, the definition of skills, and even the way humans communicate with machines are all being rewritten.
The future of software will belong to those who can think in systems, not syntax, said Sanjay Koppikar, chief product officer at EvoluteIQ, in an exclusive conversation with AIM. Koppikar has watched automation evolve from rule-based scripts to autonomous digital agents.
When Software Becomes Service
For decades, enterprise software was defined by silos. One app managed HR, another handled finance, and a third took care of customers. The best any company could do was integrate them, a patchwork of applications trying to behave as one.
But that construct is now under question. “When the transformation is happening with an agentic approach, the construct of one app trying to serve you for one particular construct actually goes out of the window,” Koppikar said.
He believes that was the real meaning behind Satya Nadella’s statement. “It’s not that something like Zoho or Salesforce or some of these SaaS products are going to just die off.”
“The idea was that the software, as we saw it traditionally… everything became app-oriented.”
He explained that the idea is fading fast. An agent can comprehend natural language requests and independently execute multi-step tasks.
For example, to provide data on the top 10 customers and their pending invoices, an agent can automatically access and retrieve necessary information from systems like Salesforce (for customer data) and ERP (for outstanding invoices), and present the findings.
This capability eliminates the traditional, time-consuming software development cycle, where a software engineer is needed to understand requirements, create a detailed requirement document, receive approval, and then spend months on development before delivering a final dashboard. This entire manual process is now obsolete.
For him, this isn’t just about faster software; it’s about architecture replacing applications. Agentic systems, he said, are moving enterprises “from software as a service to service as software.”
In a previous AIM report, it seems evident that AI is transforming SaaS into “Service-as-a-Software,” where AI agents deliver outcomes instead of tools. Indian IT firms like Infosys and TCS are leading this shift, building AI agents that automate client operations. As AI replaces human roles, SaaS evolves into agentic systems powering a trillion-dollar services economy.
The New Language of Machines
The invisible software layer also has implications for those who build and maintain it. For decades, programming languages such as Java, .NET, or Python have acted as the intermediary between humans and machines.
“What is Java, .NET, C-Sharp, Python, all of these?” Koppikar answered next. “Computer languages. Now, are they actually computer languages? See, computers only understand zeros and ones.”
“They were the intermediary languages invented by humans so that they could communicate better.”
The change, he argued, is that natural language is becoming the interface. “If English becomes that language, do you need these intermediaries? No. Now what you need is a better communication skill.”
He highlighted the rise of “prompt engineers,” saying that the creation of jargon appears to be an attempt by some trainers to significantly overcharge for their services.
But the core idea, he said, is that people who can explain what they want clearly will get better results.
Still, he made it clear that fundamentals cannot be ignored. Relying solely on surface-level skills, like good English, is not enough to excel in software development.
While some progress is possible, achieving significant milestones still requires the expertise and deep understanding of those who specialise in the subject matter. Ignorance, in this context, serves as a barrier to genuine progress.
Jobs Won’t Disappear, They’ll Multiply in Meaning
Every era of automation has been shadowed by fears of job loss. The agentic era is no exception. But Koppikar believes history tells a different story. “Since the automation journey started, since Charles Babbage came out with the first computing machine, automation means that I need speed and I need less dependency on humans so that things can be faster,” he said.
He recalled Jensen Huang’s famous remark when asked about AI’s impact on jobs: “People are not going to lose jobs. People with AI are going to take jobs away from people without the knowledge of AI.”
That, Koppikar said, sums up the transition well, as “the nature of work starts getting transformed.”
“So the same five people you employed to manage, say, network and all of those things possibly, who were able to cater to only two customers earlier, now will be able to cater to 10 customers.”
He added that the AI wave is similar to the advent of mobile phones that led to job displacement for those involved with pagers.
A Quiet Revolution
SaaS to agentic architecture isn’t just a technological shift, but a turn for humans too. As software fades into the background, people come to the forefront — not as coders or operators, but as designers of intent.
The future of automation, as Koppikar sees it, isn’t about building machines that replace humans. It’s about building systems that make humans more consequential.
With a storyteller’s flourish, he concluded: “Stories come in all forms. Technology stories become products. Non-technology stories become novels that can be printed and spoken about.”
The post The SaaS and Human Shift in the Age of Agentic Automation appeared first on Analytics India Magazine.
