3087: Web Summit 2024: Phenom - The Future of HR Tech, AI, and Regulation

[00:00:03] At this year's Web Summit in Lisbon, my guest Mahe today has been sharing his insights on navigating the increasingly complex landscape of AI regulations, including things like the National Privacy Bill and the EU AI Act.

[00:00:23] So I bumped into him on the show floor here at Web Summit and I invited him to join me and talk about some of the challenges and opportunities of implementing Gen AI in an ethical manner.

[00:00:35] And also how organizations can harness the power of AI to boost things like employee efficiency, productivity, and satisfaction.

[00:00:45] And I'm also going to ask him to look into my virtual crystal ball and look at what he believes the future holds for AI in HR.

[00:00:54] So how can companies strike that right balance between innovation and regulation?

[00:01:00] That is the meaty topic we're going to discuss today.

[00:01:04] But enough from me. Let's get my guest on the podcast now.

[00:01:08] So a massive warm welcome to the show. Can you tell everyone listening a little about who you are and what you do?

[00:01:14] Yeah. Mahi Bhairadi is CEO and co-founder of Phenom.

[00:01:18] What Phenom does is we are an HR technology company, predominantly enterprise companies, user sales, companies which are more than 5,000 employees in general.

[00:01:30] They do three things on our platform, talent acquisition, talent management, and talent growth.

[00:01:35] And we provide an infrastructure to make that happen through intelligence, automation, and what we call as an experience.

[00:01:44] We wrap the experience with intelligence and automation based on what your company's needs are, where your growth is, where you're shrinking.

[00:01:52] Okay. All that infrastructure is what we provide.

[00:01:54] We call our platform as intelligent talent experience product.

[00:01:58] And that's what we sell.

[00:02:00] Like major companies like DP DHL, like let's take an example, like ABP or Roche or Kunanagal or Regions Bank, like companies of that sort use our product on a global scale.

[00:02:15] And your work at Phenom has been pioneering in AI-powered talent experiences.

[00:02:21] And of course, here at Web Summit, AI is a huge topic.

[00:02:24] And with the recent wave of new regulations such as the National Privacy Bill or the EU AI Act, I'm curious, how do you see the regulatory landscape shaping the future of AI in HR?

[00:02:36] So the regulatory framework actually will encourage some parts of AI and discourage some parts of AI if it is actually really having an impact on a human experience in general.

[00:02:48] So we really look at the overall spectrum in six different zones, level one, level two, level three, level four, and level five in intelligence, level zero to level five in automation.

[00:03:01] When you're really operating on level zero to level two, you don't need like AI regulations won't really make a big impact.

[00:03:09] When you cross level two to cross to level three or level four or level five, that's where AI regulation will really come into picture.

[00:03:17] The reason is you're really using a much more deeper infrastructure to make the AI work.

[00:03:23] In the initial phases, what happens is you basically use scheduling.

[00:03:27] There is no human interaction.

[00:03:29] It's about the systems which are really making it.

[00:03:31] How can you simply operate upon it?

[00:03:33] Or you can really do a particular process of specifically fit scores.

[00:03:39] Like how do you really see a fit, but don't make a decision of going to be higher.

[00:03:43] So like which level you are operating, what regulation really hinders what particular shape is what is very important.

[00:03:52] And as long as you really, by country by country, it changes dramatically.

[00:03:57] And location by location also, it changes dramatically.

[00:04:00] So based on that, we have to adjust which company needs what.

[00:04:04] But most of the companies today are global.

[00:04:06] So what's happening is in one country, they might really follow one particular policy.

[00:04:12] But in a different country, they follow a different policy.

[00:04:15] The reason for that is because the way you recruit, the way you retain these countries are entirely different.

[00:04:20] Let's give an example.

[00:04:22] If you take in Europe, you don't fire people that often.

[00:04:25] If you take North America, people are fired quite often.

[00:04:28] And that is a policy of the infrastructure of the government.

[00:04:34] And because of that, how you use AI differs.

[00:04:37] And how you really provide restrictions are different.

[00:04:40] So like now, even in the U.S., if you really look at state by state, there are certain regulations which are coming,

[00:04:46] like California versus New York versus what the other countries or other states are really doing.

[00:04:51] So all these are very interesting landscape, but we are in very, very early stages of really shaping the AI regulation and safety in a very authentic format.

[00:05:00] I don't think we have clear-cut agenda yet, but it's shaping right now.

[00:05:05] Yeah.

[00:05:06] So as companies rush to adopt generative AI, especially if they're looking at boost in efficiency and productivity,

[00:05:13] what are the key ethical considerations that they should be keeping in mind, especially when implementing AI into their HR processes from hiring to employee development?

[00:05:24] Any ethical decisions around that?

[00:05:26] See, there are certain areas of tasks which you'll automate.

[00:05:30] There are certain areas of tasks which are argumentative.

[00:05:34] Yeah.

[00:05:34] There are certain areas of tasks which has to still be done manual.

[00:05:39] And what needs to be really moved to automation, what needs to be moved to augmentation, what needs to be moved to is all based on what kind of human experience is delivering.

[00:05:50] Is it delivering a biased experience or is it like inclusion experience and diversity and inclusion is a primary component in this whole equation?

[00:05:58] In which particular way it is actually shaping is what really governs all this infrastructure?

[00:06:04] I think from my standpoint of view and where we sit at Phenom, we constantly really look at like it will shape like the overall AI ethical standards.

[00:06:17] There is no doubt about it.

[00:06:18] But the most important thing is validation of the AI, which you call it as eval, especially in gen AI, is a very, very important spot to really look at.

[00:06:30] And the second thing is outcomes of the AI, what are the outcomes we're really getting?

[00:06:35] And like doing studies on those outcomes in almost like a psychometric analysis kind of an area is a very important factor to think.

[00:06:45] Yeah, 100% with you on that.

[00:06:48] And I know you've also mentioned that understanding AI regulations is crucial before you even begin setting any company policies.

[00:06:55] So what steps should organizations be taking to ensure that they are, yes, compliant, but still capitalizing on the benefits of AI?

[00:07:05] See, there are sort of, like right now, if you really look at like the biggest problem, what we'll have right now is productivity is a most important metric.

[00:07:14] Every company is watching.

[00:07:16] Yeah, yeah.

[00:07:16] And the reason they're watching is not because, hey, productivity is cool so that they can make money.

[00:07:21] The stock market is demanding it.

[00:07:24] And people think stock market is some unique people investing on it.

[00:07:29] Stock market is actually managed by common people who actually invest day to day of their money or even their pension funds into it and really see like, hey, where they can really go.

[00:07:39] Without productivity, profitability is not existing in today's economical environment.

[00:07:43] Because of that, everybody is really focused on how can I bring productivity up in the existing workforce.

[00:07:51] One of the techniques, the way you can really bring it up is using AI.

[00:07:54] And now, like how much you can use and how much you don't need to use and where you have to protect the employment in a much more authentic format is very important.

[00:08:06] But there are certain jobs which will disrupt more aggressively than anything.

[00:08:11] But the most important thread, what I'm thinking is we have to teach people adaptability is the most critical skill or a competency which people have to get into.

[00:08:23] If that is actually going to manifest, then people will really need whatever is automated, which is augmented before as a task, can be picked up and really move on to the next one.

[00:08:36] So learning becomes a lifelong lesson and whoever can do it quite effectively will have no problem.

[00:08:42] But if people are really stagnant in learning and don't want to learn more, they will have a tough time in coming years to come.

[00:08:51] One of the things that stood out to me about Freedom when I was doing a bit of research on you is that you aim to help a billion people find the right work.

[00:08:59] So how do you balance leveraging AI for efficiency with that need to uphold things like fairness and avoid bias in hiring and talent management?

[00:09:09] Yeah.

[00:09:09] So the bigger point, what we think is people really losing jobs with AI continuously and will never have a job is not a thesis what we believe in.

[00:09:22] What happens is there are particular tasks, particular equations which will be automated with AI and there is no doubt about it.

[00:09:30] But there will be a lot more human experience need which we are not doing today.

[00:09:35] Those technology restrictions which will open up to do in a different format.

[00:09:39] Like how customer support is being really thought about in the last 30 years is not the future customer support.

[00:09:45] But it will have a manifestation which is a much more different way.

[00:09:49] So for us, in the last 2,000 years, every time a technology came in, in whatever format, of course, the workforce has shifted significantly.

[00:10:03] But it never made, like the workforce is not required.

[00:10:07] Yeah.

[00:10:07] Like even AGI or ASI, I don't even think will really do that kind of an experience.

[00:10:14] The reason is, at the end of the day, human is the one who is asking for an experience and our expectation constantly changes in a very ridiculous format.

[00:10:24] So we have, because of that, like I don't think this productivity gain will be the next 5, 6 years and then it becomes a norm.

[00:10:32] Once it becomes a norm, like people will really pick up, like okay, my expectations have changed.

[00:10:38] Just automating is not enough.

[00:10:40] I need a human experience in the middle.

[00:10:41] So people will be employed from that particular fabric is what I think.

[00:10:45] And I think many companies are hesitant to implement AI due to concerns over privacy and compliance.

[00:10:53] So how can HR leaders navigate the complex landscape while using AI effectively to solve some of those recruitment and retention challenges?

[00:11:02] It seems somewhat of a balance.

[00:11:04] Yeah.

[00:11:04] So what's happening is really working with the IT departments, where the HR departments really working hand-in-hand with IT departments is a very crucial element.

[00:11:15] If you really look at like today, IT departments are using AI in multiple sectors.

[00:11:20] If you look at like cybersecurity or what they're really doing with sales, what they're doing with marketing, e-commerce, all these areas within their companies, even finance, they're using AI.

[00:11:33] So if that is the case, why can't they use on the other side too?

[00:11:36] Yeah.

[00:11:37] So like finding that common language of what is being used on a customer that you can also really use on talent is a very important fabric which can manifest.

[00:11:46] And we help our HR leaders to really think through it and really work in tandem with their HR IT infrastructure or the CIO, the CIO office to make sure they can get that comfort.

[00:12:00] And if they're not comfortable based on a policy, what they have will really work with which particular area they really work upon.

[00:12:07] Because at the end of the day, the gen AI to work, you need context of context of every company, every department, every business unit of an enterprise.

[00:12:17] What happens with the context is context is all about what is your thought process?

[00:12:22] What are your mental models?

[00:12:24] What are your leadership styles?

[00:12:26] Is this what you really build in the ontology?

[00:12:28] And that is what manifests with the restrictions, what access you want to give to whom at what stage and what kind of limitations you want.

[00:12:36] And that policy can be defined into the infrastructure and so that it will be safe and it will be actually specialized for each particular company in its own format.

[00:12:48] So true.

[00:12:48] And with Phenom's Gen AI solutions being used now by over 700 global companies, I'm curious, what feedback have you received from clients about the impact of AI on employee satisfaction and engagement?

[00:13:02] And were there any surprising benefits or even challenges?

[00:13:06] Yep.

[00:13:07] I'll give you a couple of things.

[00:13:09] So we actually automate a lot of interview scheduling.

[00:13:15] So we do per month about like 8 million interviews on the product without a human being involved.

[00:13:30] But after the interview is scheduled, a human being has to be involved in certain areas.

[00:13:35] In certain areas, there are video interviews which are automated.

[00:13:38] So we actually look at all the data so that you have amplified productivity almost like on average a company really saves in a year about like $20,000 just by using the stick.

[00:13:53] Right.

[00:13:53] Then there's another area where we use Gen AI specifically about like high volume hiring.

[00:14:01] There are a lot of jobs where you have to hire constantly.

[00:14:05] And in that particular case, we just automate the whole equation so that 90% of the process can be automated and the 10% is where the human really comes into picture.

[00:14:14] And that's another area.

[00:14:16] The third area is who in the company has to grow?

[00:14:20] Who is the person who is at risk to really need them?

[00:14:24] Who are actually going to like create an impact but really not getting in the last couple of years?

[00:14:30] How do you segment them?

[00:14:31] Give them a message and have an employee relationship management in a segmented format based on where they really sit.

[00:14:39] A person sitting in the same job for the last four years didn't change the job.

[00:14:43] Yeah.

[00:14:44] On average for that job, is this the right kind of a role or should we really rethink about moving the person into new roles?

[00:14:51] All this where we use Gen AI.

[00:14:53] But if you put aside on that particular thread, like how it manifests is very important.

[00:14:59] The way it manifests is we build this X plus assistant.

[00:15:05] You can really ask a question.

[00:15:06] It can give an answer about an employee or a candidate or a process or a particular flow and really say like, hey, what do I know about this person?

[00:15:16] What can I do?

[00:15:17] We can ask anything in detail about all these people in multiple layers.

[00:15:22] And that's something what we do because we actually constantly look at not only what is sitting in our system, we look at HCM, we look at your SharePoint, we look at your docs, your contracts, your resumes, and really consolidate the data and manifest in a different format.

[00:15:39] So that it's very simplified for a manager or a HR leader to really make the choices right.

[00:15:45] That's one.

[00:15:46] The second one is we do enterprise search very differently.

[00:15:51] Like how HR and the managers has to search for something they want, not on a conversation basis, but it's more from a particular context basis.

[00:16:02] That is also something what we do in an aggressive format.

[00:16:05] And that actually really produce a lot of results for our customers where we get about like 2 million searches per day.

[00:16:12] Really?

[00:16:13] Wow.

[00:16:14] That's phenomenal.

[00:16:15] And if we were to zoom out for a moment and look at the broader trends in AI adoption across industries, what do you think will be the most significant changes in everything from talent acquisition to employee development over the next, what, three to five years, especially as regulatory frameworks continue to evolve?

[00:16:33] Seems to be a lot of things converging here.

[00:16:34] How do you see this taking shape?

[00:16:36] Right.

[00:16:36] So there will be a lot of difference really going to happen both in talent acquisition and talent growth, which is employee evolution.

[00:16:44] But what happens is industry by industry, it changes dramatically.

[00:16:50] What does that mean?

[00:16:52] A retail industry is entirely different from a technology industry.

[00:16:56] The technology industry is entirely different from a financial industry.

[00:16:59] And a financial industry is different from the logistics.

[00:17:02] And that is different from healthcare.

[00:17:03] So specializing on how talent really moves and talent mobility actually apply within a specific industry, because those industries are also getting disrupted by Gen.AI.

[00:17:16] And each industry is disrupted differently.

[00:17:18] Is what the nuances what we're really building at a verticalization.

[00:17:22] We build each particular vertical in a very authentic format so that we can help that particular vertical and then really go into that enterprise in a much more deeper format.

[00:17:32] And that is what really gives a lot of context.

[00:17:36] So we call this context-aware experiences, ontology-driven products.

[00:17:41] So a context is at a user level, and ontology at a company level.

[00:17:46] And if you can really mix them really well, you can manifest this data in a much more beautiful way.

[00:17:52] So the experience is relevant to the people for what they're really looking at.

[00:17:56] Such a powerful message.

[00:17:58] And on a personal level, you are one of Goldman Sachs' most exceptional entrepreneurs of 2024.

[00:18:04] So I've got to ask, what advice do you have for any tech leaders or indeed business leaders looking to integrate AI into their business,

[00:18:12] while also navigating the complexities of emerging regulations and ethical concerns?

[00:18:17] I know you said as soon as this finishes, you're straight off to Zurich to speak with customers.

[00:18:22] It must be the kind of question you get a lot.

[00:18:24] But what kind of advice do you offer?

[00:18:26] It's the most important factor what people have to really think is AI should be embedded within your current workflow.

[00:18:35] You cannot really take it outside and do it.

[00:18:37] So almost like Tesla did, within the car, there is an autonomous equation.

[00:18:43] It's not outside, which is a very, very important factor.

[00:18:48] And that is what we constantly begin.

[00:18:51] Within a recruiter process, within a manager process, within an organizational development process,

[00:18:57] within what you call the talent management process,

[00:19:01] how can we really embed AI at a task level, at a process level, at a system level?

[00:19:07] It's very, very important.

[00:19:09] And by industry and the job type, you actually deploy differently.

[00:19:13] You don't deploy everything C.

[00:19:16] Because your supply-demand ratios are different.

[00:19:19] The country birth rates are different.

[00:19:21] Because of that, the population limitations are there.

[00:19:24] So I'll give an example in healthcare, if you really take.

[00:19:27] In US, we have, in the last 40 years, the lowest nurse graduates and highest nurse retirees.

[00:19:35] In the last 40 years.

[00:19:38] And the country's having the lifetime expectation, expectancy, has actually grown up.

[00:19:45] And we have lots of older people in general.

[00:19:48] But we don't have infrastructure.

[00:19:52] So in those scenarios, that's why all healthcare systems are really right now squeezed.

[00:19:57] Because of this unique process.

[00:19:59] Like that every industry has its own nuances.

[00:20:03] In those nuances, how can you use AI?

[00:20:05] And how can you use the infrastructure to help them?

[00:20:08] Is what does our meaning to help a billion people find the right work?

[00:20:12] It's all not like a cliche, one single equation saying, okay, it's having jobs, it's not having jobs.

[00:20:18] There are certain spots, the jobs do exist, but there are no people.

[00:20:22] In certain spots, jobs don't exist, but people do exist.

[00:20:25] So how do you balance all these equations?

[00:20:28] And which ones really?

[00:20:29] And how do you help people to migrate to spots where jobs are ample, but people doesn't exist?

[00:20:37] That navigation of skill morphing and skills really mobilizing is a very, very important job for our purpose to be realized.

[00:20:45] So it's not just acquisition, talent acquisitions, also talent management and talent growth is part of how we really think about the overall equation to evolve in the next 10 years.

[00:20:56] Wow, so much food for Thor there.

[00:20:58] And for anyone listening wanting to carry on this conversation, maybe they want to connect with you or ask your team something.

[00:21:03] Or just learn more about Phenom.

[00:21:05] Where would you like to point everyone listening?

[00:21:07] So Phenom has a bunch of podcasts, which we really do in general.

[00:21:12] And Phenom also do what we call as Am Phenom Conference.

[00:21:15] We do in US and Europe, which is also like a largest conference what we normally run.

[00:21:21] Those are the other avenues.

[00:21:22] And I'll give you a couple of links to all the podcasts what we really run as a company so that you can really connect, like give the link so that people can connect there.

[00:21:32] Absolutely.

[00:21:33] I'll make sure I'll at least do everything that you give me there so people can listen and interact as well with you.

[00:21:39] And I'd love to stay in touch with you and see how this evolves and maybe meet again on the show floor of another tech conference.

[00:21:47] And I know you've quite literally got to death to catch a plane.

[00:21:50] So thank you so much for stopping by and talking with me today.

[00:21:53] I think this conversation on the show floor here at Web Summit has highlighted the incredible potential of Gen AI in transforming the talent experience while also underscoring the importance of navigating the evolving regulatory landscape with great care.

[00:22:11] And my guest's vision for AI in HR is both inspiring but also pragmatic, focusing on harnessing technology to empower employees and streamline talent management processes.

[00:22:25] But as we embrace these innovations, we must also consider the ethical implications and the regulatory frameworks that are quickly taking shape.

[00:22:35] So as we look at the key questions remain, how can companies leverage the benefits of AI while also remaining compliant and protecting the privacy of their employees?

[00:22:46] I'd love to hear your thoughts on this pressing topic.

[00:22:50] So big quick question for you all.

[00:22:51] Are we ready to fully integrate Gen AI into our workplaces?

[00:22:56] Please share your insights by emailing me techblogwriteroutlook.com, neilcqs on LinkedIn, X and Instagram.

[00:23:03] Share your insights and also join me again tomorrow for another thought-provoking conversation live from the Web Summit 2024 show floor.

[00:23:13] But enough talking for me.

[00:23:15] Time for me to work that room again.

[00:23:17] Speak with you all tomorrow.

[00:23:19] Bye for now.