No Five-Year Plan

Most people are just now finding their footing in artificial intelligence. For the better part of two decades, Jeegar Shah (MBA '13) has walked all its interior pathways. Shah has been building in this space since before it had a name, long before it permeated industries and personal lives with the tools and technologies that have become impossible to ignore.

In the mid-2000s, Shah was a hardware engineer at AMD, leading its branch prediction team and doing something that was, at the time, largely unheard of: exploring how to build a neural network directly into hardware. AI was not yet a buzzword. He was curious about it anyway. That curiosity set the direction for everything that followed.

Now the Head of Applied AI and Platform Engineering at Atomicwork, Shah is still in that same territory, asking the same big question: what does it look like when AI and humans work together? Only now, the rest of the world is asking it too.

Breaking Out of the Shell

Let's work backwards. Before Atomicwork, before McKinsey and a direct involvement in four startups, came a Booth education.

Shah arrived at the MBA program as someone deep in the tech world. He knew how to build, but was unfamiliar with the business elements associated with creating something: the users, the market, the story, the business case. "I was largely a coder / builder," he says, "more focused on engineering without thinking about business outcomes or what it means for the end user."

Booth was fundamental in changing that. He spent time at the Polsky Center for Entrepreneurship and Innovation, took every entrepreneurship course he could find, and immersed himself in the kinds of conversations that happen outside of a codebase and circuit designs. In 2012, his team took home the top prize at the New Venture Challenge with Mouse House, an early sign that the technical foundation he had was being refined into something sharper.

He credits a few professors in particular with a lot of that sharpening. Learning to read a term sheet, understanding the nuances behind why certain terms and conditions are put in place, getting comfortable with the softer skills of negotiating and presenting; it all added up. "I have to thank every single class I sat in on with Professors Steven Kaplan, Waverley Deutsch and Scott Meadow" he says. Professor Kaplan, co-founder of the entrepreneurship program at Booth, Professor Deutsch, Adjunct Professor of Entrepreneurship and Professor Meadow, Clinical Professor of Entrepreneurship, were teaching a set of skills that extended far beyond any one deal structure: the ability to see the ecosystem around a product, not just the product itself.

His second year, Shah was commuting from California to Chicago each week, dividing his time between the full-time MBA program and a startup gaining traction and funding on the West Coast. In hindsight, he feels the absence of the socialization and connection-building central to the Booth experience. But with investors and customers watching, it did not feel optional. What it did give him was a real-world entry point into the entrepreneurial process, a place to put into practice what he was learning in real time.

He spent time with three more startups after he graduated, all in AI and ML, across life sciences, industrials, and enterprise applications. A pattern rooted in applied AI was already forming, even if the destination was not yet visible.

The Long Way Around

If you took a look at Shah's resume without context, you might be left wondering what the plan was. Physics. Hardware engineering. Software. AI. A handful of startups. Three years at McKinsey. Four and a half years at Amazon. A year and a half at ServiceNow. And now, a startup again.

There was no plan. That, Shah would tell you, was exactly the point.

After one of his startup exits, he found himself at a crossroads and chose McKinsey for the opportunity to polish and incorporate what he had learned at Booth. He was still grappling with the question of how to stop thinking strictly like an engineer and start thinking bigger picture. Booth had started that process. McKinsey solidified it. The two together gave him what neither could have done alone: technical depth paired with the ability to step back and see the broader picture.

Engineering leadership roles at Amazon and then ServiceNow followed, where he focused on building organizations, aligning roadmaps with what customers actually needed, and learning how to lead a team toward a shared goal. He was still technical. Just no longer only technical.

Then the startup itch came back.

That is how he found himself at Atomicwork, once again rolling up his sleeves and building within a smaller team. A novel approach to enterprise service management, Atomicwork is the platform handling the invisible friction of organizational life. Onboarding a new employee, ordering a laptop, submitting a PTO request, getting access to a piece of software. The tasks that, in most companies today, require a human somewhere in the chain. The idea is that they do not have to.

Shah and his team are building AI agents that can understand employee intent and act on it, completing tasks and resolving requests without anyone needing to file a ticket and waiting on a response. "Work should seamlessly pass from humans to AI agents and back," he says.

Which raises the question at the center of everything Atomicwork is working through: where does the human end?

Shah does not have a clean answer, and he is honest about that. Agents can be given instructions, personalities, and a defined budget of resources to work within. What they cannot be given, without care, is too much agency. "You don't want to give a lot of agency to these agents to do things that could be detrimental to society," he says. Figuring out where agents fit and what limits they need is what he calls "the artsy part of this."

And then there is the sheer velocity of the field. AI models are advancing faster than most companies can integrate them, and Shah is no stranger to that whiplash. He returned from spring break last year to discover the ground had moved again. New releases, new capabilities, new edge cases to evaluate. "You feel behind every time you take some time off," he says. For someone who has been in this space longer than most, that pressure is real. But part of the responsibility that comes with the work is keeping up with it.

A Better Question

Shah has advice for students who are currently finding their way at Booth, but leads with a disclaimer: his path was random, hard to follow and make sense of. "Do what I say, not what I do." But if he had to do it again, he says, he would do it exactly the same way.

The thing he would change is the question. Specifically, the one asked in almost every MBA interview: Where do you see yourself in five years?

Shah's answer: that question should not be asked.

"In this day and age, when we don't know what's happening in the next month, much less in the next five years," he explains, "that question serves only as a block to your curiosity." It implies a destination. It asks you to commit to a direction before you have had a chance to see what is out there. For someone whose career has moved from physics to hardware to AI to consulting to big tech and back to startups, the five-year plan would have been more of a cage than a map.

What he recommends instead is much simpler: focus on what you want to learn today. Build something, even if it goes nowhere. Talk to people who think differently than you do. Stay curious. "Even if nothing tangible comes from it, it's not a waste of a day," he says.

The breadth you gain from that kind of learning matters more than it might seem. When you are in a room pitching to investors, what they are looking for is someone who has thought holistically, who knows what they do not know and can speak to why this is the problem they chose. Breadth, in that space, reads as credibility.

What Shah's own trajectory shows is that the dots do eventually connect. The device physics led to hardware. The hardware led to AI. The AI led to software, The AI/software led to the startups. The startups led to McKinsey. And all of it, together, led to Atomicwork and the question he has been building toward for more than twenty years: what happens when humans and machines work side by side?

He got there by following his curiosity, one day at a time. Not by planning it.

"Let your curious self choose the path," he says. Coming from someone who had a foot in AI before most people knew to look for it, that is not a platitude. It is a credible track record.