The transformation of work in the age of artificial intelligence — what UiPath’s CMO has to say about it
Work as we know it is changing — and fast. Every CEO is talking about artificial intelligence, and most employees are wondering what that means for their paycheck at the end of the month.
Five years ago, UiPath went public with a pretty straightforward promise: automate repetitive tasks and give people their time back. The pitch was simple but powerful — take those mechanical, repetitive activities that drain energy without adding real value to the business out of human hands. And it worked really well. The company grew, captured market share, and established itself as the category leader in robotic process automation, or RPA.
But the corporate world evolved, artificial intelligence entered the picture in a big way, and now the conversation has gotten a lot more complex and, at the same time, a lot more interesting. What used to be a discussion about operational efficiency turned into a conversation about the future of work itself — about how companies will reorganize, which skills will matter, and how humans and machines will share the same productive environment.
Today, the company no longer just talks about automation — it talks about agentic business orchestration, about getting AI agents, robots, and humans working together in a coordinated way, like parts of a well-tuned machine. This shift in language isn’t just cosmetic. It reflects a deep transformation in UiPath’s philosophy and in what the company sees as the next big step for technology applied to business.
UiPath’s CMO, Michael Atalla, gave an exclusive interview to The Rundown AI, and the conversation was straight to the point. Among the topics covered:
- What has changed and what has stayed the same in the five years since the IPO
- Why most artificial intelligence projects still fail before leaving the pilot stage
- What is really happening with jobs
- And where the human factor still is — and will continue to be — irreplaceable
If you work in tech or follow the AI race in enterprise closely, this is exactly the kind of conversation worth paying attention to. 👇
Five years after the IPO: what changed and what stayed the same
In 2020, UiPath rang the stock exchange bell as the category leader in robotic process automation. Back then, the promise was clear: automate the task, free the person. And that worked — it still works, actually. But the landscape has changed a lot.
According to Atalla, if you walk into most large enterprises today, you’ll find dozens of automations running in parallel with no real way to connect them to each other — or to connect them to what the business is actually trying to achieve. The question clients were asking five years ago was can we automate this?. Now, the question has become how do we make all of this work together?
The answer, in UiPath’s view, is orchestration. AI agents, automations, people, and systems operating end to end, with visibility across the entire flow. It’s no longer just about automating one step — it’s about making sure all the steps talk to each other in an intelligent and governed way.
But there’s something that, according to the CMO, hasn’t changed at all in these five years: technology should remove friction from people’s work, not add new kinds of friction. That’s the bet that has remained intact since day one.
Lessons from someone who already lived through a revolution
Before joining UiPath, Atalla spent 15 years at Microsoft, leading marketing for Office during the migration from on-premise to Office 365 in the cloud. And he remembers that experience with a clarity that helps make sense of the current AI moment.
In 2011, he was demoing features like Exchange conversation view and the don’t-reply-all button for clients who were terrified of taking their email off a server they could physically touch. The lesson that stuck? You can have the right product and still lose the customer if you can’t help them rethink how work actually gets done.
During the cloud era, Microsoft wasn’t selling software — it was asking people to change how they collaborated, where they stored information, and whether they trusted an invisible system. The companies that got stuck in the cloud transition didn’t lack ambition. They simply took what they had and threw it into the cloud without redesigning anything.
Receive the best innovation content in your email.
All the news, tips, trends, and resources you're looking for, delivered to your inbox.
By subscribing to the newsletter, you agree to receive communications from Método Viral. We are committed to always protecting and respecting your privacy.
And according to Atalla, that same pattern is repeating now with artificial intelligence. The conversations about AI fixate on the model — on what it can do in theory. But companies don’t care about theory. They want to know if the thing works reliably, within real workflows, with real accountability.
The takeaway worth sitting with: if your team is evaluating AI tools right now, the question for your next vendor meeting needs to change. Instead of asking what does this model do?, the question should be what does our workflow need to look like for this to actually work? Get that wrong and you might end up among the 70 to 80% of companies that never make it past the pilot phase.
The wall most AI projects never get past
This is one of the most honest questions anyone in the industry can ask today. Artificial intelligence is everywhere — in investor decks, in product roadmaps, in shareholder communications. But when you look inside companies and ask how many of those projects actually made it to production and are delivering results, the number drops hard.
Between 70% and 80% of agentic AI initiatives never make it past the pilot phase. And Atalla is pretty direct about the reasons: AI pilots almost always run in isolation. An agent in one corner of the business, an automation in another. No visibility between them. The pilot works, leadership asks what the next step is, and nobody has a concrete answer. Costs pile up. Results become hard to measure. And eventually, someone decides it wasn’t worth it.
According to Atalla, the problem is rarely the technology itself. The bottleneck is coordination — or rather, the lack of it. Companies build AI solutions in silos, without considering how those systems will interact with existing processes, with legacy data, and most importantly, with the people who are going to use all of this day to day.
The organizations that are managing to get past this stage aren’t doing anything radical. They simply stopped treating AI agents as tools to deploy and started treating them as components of a larger, governed workflow. That’s the whole game.
When heavy investment turns into disappointment
A survey cited in the interview revealed that nearly half of organizations classify AI as a massive disappointment, despite heavy investments. And according to Atalla, nobody starts out wanting to fail at this. The ambition is there from top to bottom — from the CEO all the way to the person whose day-to-day is supposed to get easier.
When you see numbers like that, it’s almost never a motivation problem. What Atalla hears from clients is a coordination problem. Companies automated tasks, deployed AI tools, but those tools aren’t connected to what the business is actually trying to achieve. The return on investment disappears into that vacuum.
The clients who manage to break through that barrier start with a different question: not which AI tool should we buy?, but where does the work begin, where does it get handed off, where are decisions being made? Start there, and the technology choices become a lot clearer.
This is where UiPath’s orchestration pitch starts to make a lot of sense. Instead of treating AI as a standalone tool that solves everything on its own, the company argues that the real value shows up when you connect artificial intelligence agents, automation bots, and human collaborators within a coordinated flow. Each part of the system does what it does best — AI processes language and makes data-driven decisions, the bot executes structured tasks with precision, and the human handles context, exceptions, and the judgment that no machine can reliably replicate yet.
Redesigning workflows for AI matters, but the next step is making sure the tools running inside those workflows stay aligned with business objectives. When that happens, the value multiplies — each tool becomes more useful because it’s functioning as part of a system. 📈
What’s really happening with jobs
This is the part that generates the most anxiety — and also the one that suffers the most from misinformation. When automation advances, the easiest narrative is the one about destroyed jobs. But the reality that data shows, and that Atalla reinforces in the interview, is far more nuanced than the attention-grabbing headlines suggest.
The disconnect between experts and the general public
A striking data point mentioned in the conversation: three-quarters of AI experts are optimistic about the technology’s impact on jobs. But only 23% of the general public agrees. Who’s closer to reality?
According to Atalla, both groups are seeing something real — they’re just looking at different parts of the picture. The experts see what the technology is capable of. The end user sees what’s being handed over to technology and wonders what’s left for them. And that’s a reasonable reading of the signals.
But what Atalla pushes back on is the idea that human involvement becomes optional as AI gets smarter. A language model can’t ask should we do this?. It has no motivation, no sensitivity, no instinct for risk. Every system UiPath deploys still needs humans to oversee it, make judgments, and apply it in ways that generate real value. The role evolves. The need doesn’t disappear.
Is the job market anxiety justified?
The numbers are concrete: entry-level software development jobs have dropped nearly 20% since 2024, while senior positions have grown. UiPath’s own CEO has said that the company’s goal is to grow without increasing headcount. So the question hangs in the air — is the anxiety justified?
Atalla doesn’t dodge the answer: the anxiety is real and deserves to be taken seriously. A significant number of entry-level roles are being reshaped right now. And that’s no small thing, especially for people who built their career expectations based on a completely different landscape.
But the redistribution is more nuanced than the headlines suggest. Routine, structured work is being absorbed. But the work itself doesn’t disappear — it changes shape. New roles are emerging around workflow design, AI governance, and end-to-end process management. The demand exists. It’s the required skills that are different.
Atalla shared a personal point that illustrates the situation well: his daughter is 13. When she applies to college in five years, the jobs she’ll be competing for probably haven’t even been named yet. That might be cold comfort if you’re 24 right now. But it’s also not the same thing as replacement.
What UiPath is seeing in practice, inside companies that have already implemented their orchestration solutions, is a redistribution of energy. Employees who used to spend hours filling out forms, moving data between systems, and running manual validations now have room to work on more strategic analysis, customer relationships, and projects that genuinely require human intelligence. It’s not a utopia — it’s a process that has friction, requires reskilling, and doesn’t always happen fairly or quickly. But the direction is real and measurable.
The central point here is that artificial intelligence isn’t replacing humans — it’s replacing tasks. And that distinction matters a lot. A professional who understands how to work alongside AI agents, who knows when to intervene, when to delegate, and when to question an automated decision, has a completely different market value than someone who ignores this reality. UiPath calls this profile human-in-the-loop — the collaborator who wasn’t eliminated by the system, but who became the smartest link within it.
Where AI takes over — and where it doesn’t
As much as artificial intelligence advances at an impressive pace, there are dimensions of human work that remain unique — and will probably stay that way for a long time.
How it works in practice inside UiPath
Atalla gave a concrete example to illustrate. Imagine a finance team that needs to reconcile data across five different systems and chase down approvals via email. Automation handles the structured, repetitive parts — pulling data, cross-referencing records, routing requests. An AI agent steps in when there’s ambiguity — flagging an anomaly, interpreting an invoice that doesn’t fit the standard model. The person in that role stops doing the reconciliation and shifts to reviewing exceptions, making the decisions that actually require judgment.
The person’s time shifts to the things only they can do. That changes how work feels on a daily basis — which is a bigger impact than it might seem.
Autonomous agents: what’s real and what’s hype
With autonomous agent tools gaining serious traction in the market, Atalla makes a point of separating what’s real from what’s optimistic projection. According to him, the conversation about full autonomy is way ahead of what’s actually happening in practice.
What UiPath sees day to day is more specific and, in his words, more interesting. Agents are really good at handling unstructured data, making context-aware decisions within defined processes, and managing exceptions. Think document comprehension, fraud detection, and customer service triage — the kind of work where the input isn’t clean and a rules-based system either fails or needs constant babysitting.
Deterministic, rules-based work still runs better with traditional automation. And decisions that carry real accountability — approvals, escalations, anything with consequences — stay with people.
The near-term model is agents operating within orchestrated workflows. More cognitive responsibility for AI, yes. But still governed. Still observable. Still with humans at the points that matter.
While the big frontier AI players keep pushing the idea of full autonomy, UiPath’s approach is more grounded — deploying agents only where they genuinely excel and keeping humans on the higher-value work. It’s less exciting as a headline, but it keeps the machine running with concrete results. And that’s the version of AI adoption that actually holds up over the long run.
Quick-fire answers from Michael Atalla
To close out the conversation, a few direct questions that drew some pretty revealing answers:
What’s the biggest mistake companies make with AI?
Atalla: Expecting it to fix a broken process. AI makes good workflows faster and bad workflows more expensive.
What do companies get wrong when introducing AI to their teams?
Atalla: Framing AI as something happening to people, instead of something they’ll build with. The anxiety starts with that framing — and it’s usually avoidable.
If you weren’t at UiPath, what AI problem would you want to solve?
Atalla: The gap between what organizations believe AI will do for them and what it’s actually set up to do. That’s a clarity problem, not a technology problem. And it’s genuinely fascinating.
What to take away from this conversation
The interview with Michael Atalla paints a pretty grounded and honest picture of where we are with artificial intelligence in the enterprise. It’s not that futuristic talk disconnected from reality, and it’s not fearmongering about robots taking all the jobs either.
The core message is: the technology is mature enough to deliver real value, but only for those willing to rethink how work gets done — not just for whoever buys the shiniest tool on the market. Orchestration between AI, automation, and people isn’t just a UiPath product strategy. It’s increasingly the way organizations that are succeeding with AI are structuring themselves.
For anyone in the job market, the message is also clear: the skills that will matter over the next few years aren’t necessarily the same ones that mattered until yesterday. Adaptability, critical thinking, and the ability to work alongside intelligent systems are becoming real competitive advantages. And the sooner every professional internalizes that, the better positioned they’ll be to navigate the changes that are already underway. 🚀
