The comfortable lie in every talent decision

Long before I ran HR for anyone, I was an artist. There was an early exhibition of my work at the Academy of Fine Arts in Calcutta when I was fifteen. Almost no one who works with me would guess it, because they know the other half of me, the one who runs on numbers and reads a set of figures for the pattern in them, because the pattern tells me fast where the real problem is. I run on a simple rule that a stalwart told me once, in God we trust, everyone else, bring data!

I have known from the outset that these two instincts are not in tension. Math is abstraction and close to art in its best version. We all are a smorgasbord of ideas, thoughts, and instincts that help us to navigate this world, learn new concepts, and adapt to change. Conventional talent management uses the surrogate of the job experience to identify the right person for a defined job. In doing so, a reductive process of counting years in a specific type of account, process, or domain takes place.  

While that methodology has served us very well in a stable mature environment, it no longer will. We need new ways to find our brilliant people—those that learn well, are infinitely curious, and able to see patterns. That part is not entirely in the data, and must be discovered through conversations, listening, and in constant questioning of conventional decisions.  

The trap is in the hypothesis you make about talent

That is the trap sitting inside most talent decisions today, and I want to be honest that I have built systems in the past that perpetuate the convention. The problem is not a shortage of data. It is in the hypothesis that measures what is easily measurable and conflates it with real talent.  

Tenure is the cleanest example. It is easy to measure, so we lean on it. How long someone has run a system becomes a stand-in for whether they can run what comes next. At the speed things move now, that stand-in is breaking, and sometimes it is inverted. The longer someone has run today's system, the harder it can be for them to question it. Deep familiarity with a stable process can become the thing that stops you seeing that the process should change at all.

It is most likely that my high performer of yesterday is someone struggling to transition to the new world. The typical path of the last two high ratings along with a unique familiarity of enterprise culture leading to a promotion is a construct of the past. So, if performance data or tenure are not cutting it, how do I find the right people for the future?

Add to that the layers in the organization that were created when scale meant more oversight, and growth for people meant larger teams to manage. It led to a structural problem: information asymmetry, where the people closest to the work cannot act because what they see never reaches where decisions get made.  

Hierarchy that adds layers instead of speed. Too little room for people to use their own judgment and be wrong sometimes without it ending them. Those three things punish exactly the people you most want to keep, the ones with the instinct to move before they are told to. If you are serious about readiness, those are the walls you take down first.

You have navigated this before, but not this fast.

None of this is the first time the ground has moved under this industry. I have watched it happen more than once, from the early outsourcing wave to robotic process automation to where we are now with AI. Each time, the technology gets the headlines, and each time the real question underneath is the same, and it has nothing to do with the tools.  

Will our people be ready for what is next, not what is already here? Not ready for this year, but ready for the next five to ten, which is the only horizon that has ever mattered.

What is different this time is the clock. The window to get people ready used to run in years. Now the short term is a few weeks and the long term is, at most, a year. So, the leader I look for runs at two speeds at once, delivering today at full capacity while redesigning the system underneath it.  

I am betting on potential over performance, on who can meet what is coming, not only who has mastered what exists. That is a harder bet because performance is visible and potential is not, and it asks you to trust the half of a person the scorecard was never built to see.

What to look for when tenure tells you nothing

We have arrived at a set of attributes we believe are fundamental to surviving and thriving in an ever-changing environment. They are not easily trainable in the short term. They come with the person and can be built upon and augmented with the skills Firstsource will need to build an intelligent enterprise.  

We call them Meta Skills because they sit underneath the skills themselves. They are the enduring capabilities that shape how someone learns, thinks, and responds, and they travel with a person from one problem to the next, even when the context, technology, or job itself changes. They are not another checklist of technical competencies, and they are not proxies for a particular role or domain.

We think about them through four dimensions:

• The ability to make sense of what is in front of you and exercise judgment, which shows up as creativity and systems thinking, the knack for looking at a mess and seeing the shape of it.  

• The ability to learn from experience and carry it into what comes next, which is really learning agility, the instinct to change how you work with the tools already in your hands and to ask for more because you can see something worth chasing.  

• The willingness to own an outcome rather than wait for someone else to solve it, which tends to look like a low tolerance for a process that survives only because no one has questioned it since the day it was built.  

• The ability to absorb feedback, recover from setbacks, and keep improving. People who solve from first principles usually score high on all four, and they are the ones who cannot leave a broken process alone.

These are the qualities that make potential visible before experience catches up, and they become the foundation the rest is built on as the business evolves. You will not manufacture them in a two-week course. They are dispositions people bring with them, and our job is to recognize them early and build on them.

The instinct to change how you work with the tools already in your hands, to ask for more access because you can see something worth chasing. Creativity and systems thinking, the ability to look at a mess and see the shape of it. A low tolerance for a process that survives only because no one has questioned it since the day it was built. People who solve from first principles tend to be the ones who cannot leave a broken process alone.

None of that shows up on a resume, which is the entire point. It is the part of a person you have to see rather than score. Sundara and Ashish on our leadership team have already built a group of exactly these people, fresh graduates hired as forward-deployed engineers, people brought in to work right alongside the problem rather than at arm's length from it, chosen for how they think rather than what they have already done.  

That is what betting on potential looks like once it stops being a slogan and becomes a hiring decision you sign off on.

Prove it before you trust it

I do not trust an idea like this until it survives a real decision, one where I have something to lose if I am wrong. Ours came in how we assess communication for customer experience roles.

The old way ran across several separate steps. Each was reasonable alone, but together they were slow, inconsistent, and hard to stand behind at scale. So we started with rigor, not with technology. We built a shared, specific definition of what good sounds like in the role, one that every hiring manager across every location could work from.  

Only then did we train an AI tool against it, and we ran it alongside the existing process on every candidate, side by side, before we changed a single thing. I set the bar for what would count as proof before I saw who cleared it, precisely so I could not talk myself into the answer I already wanted.

On the strength of it, we made the call. For one part of our customer experience hiring, the AI now carries the full assessment, with an audit trail behind every decision, faster for the candidate and traceable for us. And here is the part I care about most. For the work where judgment is harder to standardize, the human stays in, deliberately. The data showed us where it could be trusted and, just as clearly, where it could not, and we acted on both halves of that answer, not just the flattering one.

But spotting the signal is only the easy half. The harder half is what you do once you have found it, because readiness is not something you hire for once and tick off. It is something you keep building, on two clock speeds at the same time, one that keeps people sharp in the model we run today, one that pulls them toward the very different skills the next few years will demand. Most reskilling picks whichever of the two is easier and calls it done.  

UnBound is our refusal to do that, learning that meets people in the moment of the work rather than in a course they sit through once, personalized to where each person is. It is also why we word the promise to our people the way we do: Brilliant People, Bold Tech, Uncommon Careers. You cannot ask someone to bet their career on what is coming unless you are visibly building the path they will walk to get there.

The half no measure will show you

I opened with the painting because the whole argument sits on it. The artist's eye and the analyst's mind are not opposites. They are the two things you need to see a person whole, and most of what we have built rewards only the half that fits on a page.

Somewhere in your organization right now is the person who could handle what is coming, sitting in a role you decided they had not earned yet. You did not overlook them. You had a reason each time. That is exactly what makes it so hard to see.

AI is not the thing forcing this question. It is just the thing that finally makes the reason not good enough.