AI’s resume—from an HR lens

What’s it like to be CIO today, especially in an industry that is directly facing AI’s intersections with the talent landscape?

For me, a CIO, the industry is always incidental. AI is redefining the entire IT landscape, irrespective of the domain. We are using AI on two fronts—for internal employee experience and talent experience. In the former area, we are giving our employees the best-of-the-world tools for elevating productivity and intelligent automation collaboration. We have approved LLMs with a walled-garden approach. It does not let our data get out, but offers all the benefits of an LLM. Employees can leverage AI assistants that help in various tasks, but also help with brainstorming, getting new ideas and driving productivity. It’s not just an assistant but a colleague–helping an employee in several ways. On the latter front—we can help the industry in a big need area on talent experience—which is that talent wants to be found quickly and placed in the best opportunities. Customers also want this part done quickly and precisely. Earlier, this involved a lot of manual work between onboarding a talent resource and starting the work. The time window has shrunk a lot now with what we are exploring with AI and automation.

Is there any dogfooding approach here—do you try your products first? Any lessons you have garnered in that phase? Any encounters with data-related bias?

Definitely. Some of our onboarding solutions are applied internally before being offered outside. For example, an AI tool to create job descriptions without bias—that has been improved a lot with phenomenal feedback—like how sometimes it can lean towards a specific age-group or gender or a specific type of talent or region. AI can get biased with the same set of data, and that’s where responsible AI policy and legal/privacy angles are important. We are very cautious on those fronts. AI has its issues. Bias is one among them. We are very careful when it comes to talent sourcing and identification. We are happy that the time window has come down from days to minutes, with proper guardrails on bias. We are working very hard on putting in strong guardrails and policy frameworks through the right data usage. We are always working on eliminating bias.

In 2027 we might reach that stage where we all have our own agents—and they are talking to other agents.

Are we ready for vertical LLMs in the HR industry?

Not to a great extent—from what I know. We are one of the few trying it out in this area. We are cautious because it can go either way—depending on the level of control an enterprise has.

Can technology help in a specific way in driving DEI?

I remember a personal experience with AI here. I experimented with AI to create 3 job profiles. It was a hypothetical case. I was surprised to see that it pushed me back and reminded me of some imbalance with the job-description tool—thanks to the in-built guardrails.

So, speed at the top of the talent funnel is great? But what about making the experience less dry, less boring and more engaging for the applicants?

We are planning to do that. Gamification and the ability to find a job yourself through new tools—all that is possible. It will happen in the future. Right now, engagement in post-joining phases like training, coaching, development material, etc., is happening.

Do you ever come across the Moravec paradox? What’s the walking or reading faces part here- which HR areas are still intuitive and easy for humans but tough for AI to replicate?

There are still clues in interviews that many candidates give—like nervousness—that humans pick faster than technology does. AI is still not there. For example, a simple task like sending documents to customers may sound very repetitive and structured, but when the file names change or when small variations occur in content, AI can struggle. And refers it back to humans. We have to train AI in many nuances.

Any examples of AI’s impact in a specific HR area that you can recall?

The Job Description generator has helped a lot with speed and with bias controls. Customers are also speeding up the onboarding of talent. We were able to cut down a lot of manual work into minutes. The payroll areas and expense reimbursement parts are also being accelerated with AI. We have been able to specifically use automation for repetitive and boring tasks.

Can AI help in the unpleasant areas of HR-like exit interviews, letting people go, or difficult appraisal conversations?

These have not been explored yet. I think these are areas where a human touch is inevitable.

Is a hybrid human-agentic workforce a near-term reality? Would it have humans and agents as colleagues on the same level or humans as supervisors?

Both are possible. W. Ross Ashby’s Law of Requisite Variety also jumps in here. To handle variety and complexity, we need human interventions in some applications. In some cases, humans in the loop will help, but feedback from agents can also be invaluable.

How would agents be rewarded? With reinforcement? With data? With feedback?

In ML, you can do reinforcement. LLMs belong to a predictive model, so the ‘next action’ is a good path- like telling a human being. Feedback is the only way here.

How far are we from a scenario where two agents are talking to each other instead of their human owners?

In 2027, we might reach that stage where we all have our agents—and they are talking to other agents. We can envision them sitting in AGMs and financial calls already. But this can also bring chaos and confusion unless good governance is in place—especially when they start getting more and more intelligent. An agent may read a human’s mind, what one is thinking. Humans restrict a lot of thoughts and do not convert every thought into action. An agent may not have those filters. That can be dangerous.

Are areas like AR and VR intersecting well with AI and ML?

They are still in siloes. Cutting down these siloes can add a lot of value. They can make a big difference to training budgets.

There are still clues in interviews that many candidates give—like nervousness—that humans pick faster than technology does.

Will HR now be more than a stand-up comedian’s Rangoli joke?

Our colleagues in resident HR functions are leveraging AI with a lot of effectiveness, and they are enjoying creativity in many areas. They can transform a simple meeting into something engaging and compelling. HR departments would be more equipped now to deal with human emotions and leverage gamification.

What kind of solutions are you working on next? Do you also address the workplace needs that Gen Z has?

We are trying many areas to make employee experiences really world-class. They can find solutions to repetitive tasks and feel more valued. Incidentally, a big part of these projects is driven by Gen Z teams with their mindset and articulation.

What is the future of technology and HR?

In the future, HR will be empowered by intelligent automation, freeing them to engage in truly high-value, strategic, and empathetic work. The employee experience will become hyper-personalised, continuously optimised by AI, and incredibly efficient. Ultimately, I see technology enabling us to build a more fluid, inclusive, and human-centric world of work, where the right talent can always connect with the right opportunity, and every individual can thrive.

Why are many enterprises not able to move beyond POCs?

It could be due to trust issues—what if AI gives a wrong answer? But by bringing humans strongly in the loop, with the right set of products and having a diverse team (like a responsible AI officer—, his trust can be strengthened.

pratimah@cybermedia.co.in