Episode 6

Sumeet Garg on the Future of Recruitment: AI and Human Adaptability

TIM: So, Sumit, last time you were hiring, what were the biggest challenges for you in this kind of market?

SUMEET: So in the current market, I believe the biggest challenge right now is that there are like too many candidates applying for the role. So you see like for a single job ad work that we probably place out, it's like 300, 400 CVs that just come through. And ah with generative AI kind of ah being there and accessible to everyone,

SUMEET: I think ah after just going through five or six CVs you start feeling that every CV practically looks the same because they just mirror the job description. So it becomes and it has become extremely challenging as a hiring manager and especially as a recruiter to kind of really figure out who is the right candidate, right? And just like an interesting thing. So one of the CDs that we saw at the herring itself, they mentioned not created by a charge GPT, right? So people have like kind of gone on to that extent to kind of showcase the value, you know, that we're really putting in the effort and not just like curating it via charge GPT or via generative AI.

SUMEET: right So I think that's the biggest challenge as a hiring manager right now, to figure out the right candidate. ah That's one. And secondly, ah I think this this the technical assessment part. So ah a lot of candidates just mention a lot of technical skills that they have, even if they have not really worked on those tools and technologies.

SUMEET: So even if they look great on paper in terms of their work experience and everything and when you like start interacting with them especially after the after the HR round because they're just like kind of doing basic hygiene checks and when the first technical or a hiring manager interview comes up so then it just kind of shows you know that you do not have those technical skills. So for example, in one of the interviews that I was taking, like the person mentioned that they have ah machine learning experience, they've built predictive model models, and so on. But then you ask them a simple question that, how do you test the accuracy of your model? Is your model reliable or not? Blank. So that they're not able to like articulate very well. And that's the point where you know you know that. But it's not like, ah

SUMEET: no someone who has really worked on on those aspects, right? So that becomes challenging and it's like a kind of a waste of time for the HR team, for the hiring manager. ah it It becomes very, very ah tricky from that sense.

TIM: Yeah. it's interesting. Like I feel like CVs and that screening step have always been a big challenge because anyone could have always put something on their CV. They could have described themselves as a whatever it had experienced in whatever, but now it's just. It's like, yeah, the marginal cost to apply for a role is almost zero. If you can literally use an AI tool to tailor the CV to the job ad, and then also the application bit as well. I think this.

TIM: some products in America that candidates would use to say, here, apply to these 50 jobs, even filling in the form and all those kinds of things. So this is only going to get worse before it gets better, I assume. What what do you think?

SUMEET: Yeah, absolutely. I think like at the end of the day, you want someone who can really like add value to your team with respect to everything, with with respect to the technical skills, with respect to the soft skills. ah So technical skills is still something that you know you can You can still probably learn on the job as long as you have the right attitude, you have the right soft skills. But I think like as of initial screening, it's it's becoming more and more ah challenging and discoting you from your example about like someone applying to 50 job adverts with like probably a few lines of code. And I did see someone post that actually on LinkedIn, where they said that they had done that.

SUMEET: And it's it's just ah you know fascinating to kind of see where technology is leading us. Obviously, there are like pros and cons of it. But like from this perspective, in terms of how it's being used, it becomes really challenging as as ah as a hiring manager who wants to bring in the best of the candidates, wants to bring in the best of the talent to your organization so that they can add value. It just lengthens the time process.

TIM: Yeah. it's also struck me recently in having conversations that, uh, candidates would view this problem, uh, in a similar, but slightly different lens, which is like, Oh, I've applied to all these jobs.

TIM: I can't get a callback. I need a tool that's going to allow me to apply to even more jobs. And in that mentality, they're almost creating a problem for themselves in a sense, because now surely worry, reach a point where.

SUMEET: Yeah.

TIM: the applications won't be screened by a human in the next year. Like, I think the technology surely will have to be in place to do some kind of automation with AI or any other tool, because it will be literally impossible for a company otherwise to deal with all the applicants.

TIM: And so it's interesting to hear that kind of complaint from candidates.

SUMEET: Yeah, definitely. but Yeah, I think it's it's like a very, very, it's it's kind of becoming a catch-22 kind of situation where, you know, you're just like kind of oh as a hiring manager, you're expecting something, the candidates are kind of expecting something else. You're just kind of going in loops trying to fit into like what the other person is looking for. So it's it's just becoming a bit messy out there. And I think at the end of the day, a technological ah kind of ah ecosystem will

SUMEET: be very, very helpful to kind of come in and kind of eliminate ah these instances where just like people are applying like left and right to any and every possible job that's available there.

TIM: Yep. Do you, do you imagine that, I dunno, job platforms would have to then be incentivized to reduce the chance of like an automated application? because at the moment, if I think about it, like when we've used LinkedIn jobs and those kinds of tools, if anything, they'd be promoting the opposite. They're like, Oh, we're going to maximize the number of eyeballs your ad gets in front of them. We're going to maximize your applicants. But it's sounding like that itself is almost a vanity metric that you need.

TIM: kind of almost like validated applications or kind of real applicants or or something like that. So the job systems themselves are going to have to shift up something as well. I would have thought, what do what do you reckon?

SUMEET: Yeah, absolutely. I think we we kind of need to go beyond beyond the quantitative metrics, saying that, OK, we're just giving you eyeballs to 100 applications and stuff. The real metric has to be eyeballs to 100, for example, 100 applications which suit to your needs and to your requirements and to and to your goals. And like just like from a ah resume my point of view, I think but people leveraging AI to curate their resumes is going to be more difficult. So something has to like come in place which goes beyond their resumes and evaluates them at the very first initial step itself.

TIM: Yeah. one thing I was also thinking about recently, uh, was said this whole problem of like, okay, I'm a candidate. I'm now going to use AI to apply on master jobs to tailor the CV. Oh, I'm a business. I'm getting too many applicants. I'm going to have to use AI to combat the candidates AI. And there's this kind of race to the bottom almost none of that.

TIM: exists in like the shadow job market or like the oh, I knew someone who knew someone who got me a foot in the door. Or like I used to work with a person and that's how I got the job. I just spoke to someone just last night who said they haven't applied for a job in 10 years.

TIM: Like the last three roles they've gotten was through a coffee chat from someone they knew because they're in the industry.

SUMEET: Yeah.

TIM: So

SUMEET: Yeah.

TIM: Do you imagine almost there's going to be suddenly this value to the network side of things or to who you know even more than they used to be as well, perhaps?

SUMEET: oh I think that it's happening even today and like with with the current job situation I think you you kind of read posts on LinkedIn every day in terms of of recruiters mentioning that you need to like network a lot, ah you need to attend events, you need to attend probably job fairs and all that thing. So I think networking has always been there, ah to be honest. like I have experienced that. I have been on the side, on the advantageous side of it as well, on the disadvantage side of it as well. But I think at the same time, in in modern day times, it's been a bit more aggressive just because there's like a lot of imbalance in terms of the job market right now. There's like so many candidates, but there's like so few jobs. I think I was reading a post on LinkedIn which mentioned that

SUMEET: ah People with like an open to work tag on their profiles. That's like probably ah one APK or something on LinkedIn. ah Whereas the active jobs that are there on LinkedIn is probably 150 K or something.

SUMEET: So there's like a huge imbalance in that sense.

TIM: Yep.

SUMEET: And this is where like networking with the recruiters with potential, ah hiring managers, sending out them requests, sending them like ah your CVs directly to them, onlying them via email. So that's been happening quite a lot. I have received quite a lot of messages ah through that approach. I have tried to do the same in my present ah kind of in my in my present situation of looking for a new job. So I think that's it's it's

SUMEET: always been there but now it's just like ballooning given the current situation.

TIM: Yeah.

TIM: What about last time you were hiring for a data role data role, as people would say probably everywhere in the world except in Australia?

SUMEET: yeah

TIM: What was your hiring process? Like what were the steps you went through and why did you set it up that way?

SUMEET: Yeah, so ah I think the key starts off with the idea of having a nailed down job description. So as a hiring manager, you should really know what you are looking for, ah what skill sets do you need, which ah which kind of add value to your organization. Nailing down that job description is crucial. From there on, you kind of start with your initial screening by the recruiter, either to LinkedIn or to other platforms, you advertise the job and then that

SUMEET: ah Process starts to recruit a kind of screens and skims through all the CVS they have received. They probably shortlist a few. They kind of have that initial round to understand basic hygiene questions, their fitment in the company, salary expectations and everything.

SUMEET: And then the second round typically that we do is like a technical round. ah We usually try to do it like face to face, checking on their technical abilities, technical skills. And this is where I think is like the biggest drop in terms of candidates moving forward. So for example, if we have like 50 people just destroying in a number there,

SUMEET: that have been shortlisted by the recruiter for like a first technical round. Probably we just kind of skim that down to five or eight for a further interview with like the senior ah managers and the SVPs within the team. I think ah and it's not because we want to bring that down, number down significantly to five or eight.

SUMEET: But it's it's just that the skill set is not there and it does not reflect ah based on what they have mentioned in the and in the CV or in the resume. So a lot of people technically they just like are not up to the mark. And this is like the biggest drop. I think this is also relevant because that's very probably as a hiring manager you end up like kind of wasting a lot of time.

SUMEET: at your end because you're just kind of scheduling 30 to 45 minutes with each of these ah ah kind of candidates and it's just kind of ah not leading you anywhere.

SUMEET: So it's like kind of finding a needle in the haystack of like a bunch of candidates.

TIM: But.

SUMEET: So we tried to kind of introduce ways wherein, you know, ah we started giving out a technical assessment ah before ah the second round of interviews ah and for them to kind of complete it within a day, depending on the role that we are hiring for. for And the analyst was different for someone like a team lead, it was different just to kind of assess their abilities.

SUMEET: It helped because a lot of candidates never like came back with the assessments. So it just kind of reflected you know that they were probably not interested. They felt that probably it was not their cup of tea given given the ah Skills that they were being assessed at but I think now with chart GPT and Gen AI We do see like that changing because people can just like leverage those platforms to provide ah Responses to those assessments so yeah a bit of insight into that and post that we usually just have like a Manage around or like around with like one of the VPs or the SVPs to kind of just check for cultural fitment and and

SUMEET: Post that, we try to close the process. So typically ah three to four rounds, and the overall process is something around ah three to four weeks usually. So we try to close a role within these timelines and like within these number of rounds.

TIM: Yeah. it's interesting. So you mentioned the biggest drop off is basically candidates have been shortlisted who on paper look good. HR team says, okay, these look good. You think from your more expert perspective, they look good. Then you interview them and they don't have the technical skills they claim to have in your experience. Do you find that it's, ,

TIM: that it's like a complete lie like they put on the CV. Oh, I've experienced machine learning and they're like, what's machine learning? Or is it more like an exaggeration? of The truth is that I have advanced SQL skills, but their perception of advanced is just wrong. And they only know basic things like what have you typically found is the case? Like what's the the gap?

SUMEET: oh So the major gap has been in terms of what a candidate feels is like, as you mentioned, like that they are proficient in versus how we as a company kind of internally identify what proficiency in a particular technical tool is. So I think it's it's that gap because there are a lot of certifications available right now.

SUMEET: that people can kind of do for technical skills. There's Coursera, there's like these platforms which provide those kind of courses. But I think that the real challenge comes is that like people just tend to kind of believe that having done that course is like they have that mastery and the expertise within that tool which is ideally not the case and I've like seen some really good candidates who have done that who do not have like a live experience or working or using a tool in like real world experiences or or like and on on on a job experience

SUMEET: But like they have leverage platforms like Kaggle, for example, ah to kind of hone their skills, to kind of work beyond their regular roles to kind of practice a lot. So I think that's what is an added thing. But yeah like to to just answer your question in a few words, it's mostly ah how they believe and see their skill set versus how we see. And there's like a gap in there.

TIM: Yeah, that's really interesting. On Aluba, we have in like the testing process before the candidate starts a test, we directly ask them, Hey, on a scale of one to 10, how good do you think you are in XYZ skills that they're about to be assessed in? Then once the test server, we give them like a self-awareness chart, which compares their perceived ability to their actual ability.

TIM: And to cut a long story short in five years of data, the average person drastically over over-estimates their skills. It's just almost like a inbuilt human bias or something that we have where we, we think we're a bit better than average.

TIM: I've heard of those surveys where it's like, oh, 90% of people think they're above average drivers. It must just be kind of like a built-in bias we have, or maybe even people are, maybe it's like an inflation thing as well, where it's like, if you were brutally honest on your CV,

SUMEET: Yeah.

TIM: if no one else is being honest, uh, you wouldn't ever get a call back.

SUMEET: yeah

TIM: Cause you're like, Oh, I have basic skills in X, Y, Z and kind of I'm okay at this. And I'm just beginning here. Then the person reviewing it would almost have to take you at face value. So it's ah there's like a built-in inflation in the whole process or something.

TIM: I'm not sure. Can you think of like an origin of why this might be happening?

SUMEET: oh I do agree with you that there's like a built-in inflation in terms of how everyone kind of projects and presents ah their skills. And like, I think we as hiring managers, I think we take that with obviously a pinch of salt because we know based on our like prior experience as well that when you actually get on with a role or with a job,

SUMEET: we we fully understand that they you would not be like knowing things 100%, for example, for SQL.

TIM: Thanks for watching.

SUMEET: So you would be probably exceptionally writing at ah create code. You have like a strong ah ability to kind of ah extract data, present findings, but like you may not be optimizing your queries very well, for example.

SUMEET: right So you're you're writing ah a hundred lines of code when it could have been done in like probably 10 lines, right? So I think, and this is something which can only kind of develop when you like constantly ah work on those technical tools. Like for example, if I am going to work on a technical tool now for any query, I'll probably write a code which is probably 50 lines because that's how I have written it. Whereas someone who has been more hands-on on a daily basis, they'll probably do it in like five lines worth of code. so just giving an example. So the more you work on it, probably the better you get on it. But yeah, like the the inflation bias is is definitely there. And we, at least I personally you always factor that in, that if you're expecting someone to be a 10, or they are saying that they are a 10, there will definitely be a seven or eight always.

TIM: When I, excuse me, used to interview candidates, I used to also like, preface it by asking a similar question like how how good do you rate yourself in SQL?

TIM: If a candidate ever said 10 to me, I'd be like, okay, let's go for it then.

SUMEET: Yeah.

TIM: Because you've almost set yourself up to get absolutely hammered. And if you don't really knock those first few easy questions out of the park, I think it says a lot about your self awareness and coachability.

TIM: That's the feedback I've heard from other hiring managers who would hear that really high perceived skill from a candidate.

SUMEET: Yeah.

TIM: Does that send a red flag for you as well of a candidates like oh I'm 10 out of 10. I'm almost unbeatable in this skill.

SUMEET: Yeah, yeah, absolutely. And typically when we are doing those technical assessments like during an interview round, so we start off with like probably a more difficult question and then slowly go down towards the easiest ones just to kind of understand and probably, and I personally feel that that's worked well.

TIM: Ah, interesting.

SUMEET: And we are able to see that, okay, probably the most difficult one, they were not very confident. But like after that, the more medium ones or the easy ones, they were able to do it. So it's just like a reflection of you know their real, where do where do they like actually stand in terms of their proficiency for the tool.

TIM: Yeah. I'm thinking about it now. I wonder whether, so CVs have always had the issue of, okay, people might exaggerate or lie, or they might have perfect self-awareness over their skills. I'm guessing this is going to be exacerbated then if most CVs soon are written at least partly by a large language model, because there won't be any guilt or like association with someone sitting in that they're writing, I have five years experience in x when they know they don't because now it's kind of like, they've outsourced that tool to the LLM, they just get back an output of a CV that they haven't written themselves.

TIM: It's almost like it's disconnected from them. They haven't committed the lie. If you see what I mean, like, I wonder whether this problem is just going to skyrocket in terms of the difference between reality and the presentation of the candidate.

SUMEET: Yeah.

SUMEET: Yeah, I think absolutely. I think it's. also down to the fact in my opinion that that's just how the job market is right now and then there's no real not really a control in terms of like how AI is being leveraged like at the end of the day if like somehow there is a tool which kind of is able to figure out their actual technical skill set right before even like you are putting your eyes on their resume. resume I think it it will just help to kind of ah eliminate a lot of these resumes and if every organization starts to practice it candidates are I think very very smart they'll get to know that okay this is happening so we cannot like

SUMEET: keep on lying or keep on exaggerating on my ah resume. resumee So they they kind of balance it out. And I think eventually this is a direction where we are moving to. We see a lot of organizations have started using those psychometric tests, using those behavioral tests, using those technical assessments, like even before a hiring manager or a recruiter can put their eyes on your CV. Right. So I think that's probably going to be the norm eventually.

SUMEET: yeah And that is probably also the right way to go about it. Obviously, some people will come back and say that there are pitfalls with this approach as well because ah the For example, when you're evaluating for technical skill sets, youre you're kind of throwing a basic syntax-based question to someone who is being hired for a head of analytics role. so It's the the questions, the abilities, the technical questions that need to be catered to the role that you're hiring for. And I think eventually everyone would get there in terms of ah their recruitment and their hiring process. But yeah, that has to be the way forward.

TIM: Yeah, 100%. So we've talked about AI and its impact, certainly in those early application stages where candidates are using it to craft a quote unquote perfect CV, where they're potentially applying, applying, sorry, en masse to different roles. Now the companies are definitely using different versions of chat TPT of different versions of AI to do that kind of initial filtering ranking of applications. So that's happening in those early stages. I've also seen some kind of like AI interviewers come out where it's, you know, you're interviewing with ah an AI machine, not a human. Do you have any other sense of how AI is going to impact hiring in the next couple of years?

SUMEET: you I think definitely yes. I think there would potentially be a push towards ah video or audio based hiring and that's something that like I've evaluated in one of my previous roles ah where we kind of tried to ah introduce this like video-based responses to questions just to understand the body language of the person because the role was like a client-facing role and we absolutely wanted to show that absolutely wanted to be sure that like they were articulate in terms of their ah communication they were able to communicate stuff properly ah they were able to kind of explain technical things in a non-technical way for example so that's something that's ah that's helped and I'm pretty sure that like

SUMEET: ah with the AI coming into the picture. So that would be something that would like ah be at the forefront for a lot of companies. ah I think even when it comes to hiring, so like headhunters is like a major thing when it comes to like hiring at senior levels. Like I believe AI can be leveraged for that as well. So just like like currently a headhunter would like just go through LinkedIn profiles and see if like who was suitable for a role. I think AI could definitely replicate that by just crawling through LinkedIn and evaluating, you know, if and coming back with a list of 10 or 15 people, you know, that you should definitely reach out to them rather than like doing that process manually.

SUMEET: So I think that's going to be a big thing and probably the next big thing when it comes to headhunters based hiring. Testing for EQ and not just IQ, I think that's going to be crucial. A lot of companies already do psychometric testing, behavioral based questions, Amazon does that. So that is going to be crucial going forward as well.

TIM: Yeah, I feel like recruitment is ripe for this because there's just so many manual steps, like a traditional recruitment process is almost 100% manual. And a lot of the work is catastrophically tedious, especially in those early stages of like, I got to write this job out, I've got to come up with a LinkedIn search or whatever search to refine down to this set of candidates to do some kind of outbound, I've got to read these hundreds of CVs and stack rank them, like, all that stuff is just primed surely to be automated away. And this is for the best, I think.

SUMEET: Absolutely and I think and I've seen that you know you at times miss out on a lot of good candidates because of this manual process like I was like recently reading a post in LinkedIn where like the hiring manager put in like a dummy CV for a role that he was going to hire but like probably ah he he kind of crafted the CV to the best of his abilities because he knew what he was hiring for and that CV didn't get shortlisted by the recruiter right so Just to eliminate these subjective biases and in today's market like for example, I don't think and I've seen that like a recruiter would not glance to a CV for more than 10 to 15 seconds. So it's so subjective now.

SUMEET: And this this needs to be ah more objective, right? Because like, thinking from that perspective, like a recruiter, typically when glancing through a CV, he's just like doing basic hygiene checks, are the right keywords there, ah ah is the right experience there, has that... ah the number of years that they work with is like in line with what the expectations are and so on. So all of these processes can be automated using a AI-driven tool which can evaluate all of this, right? And then that makes the job easy practically for everyone who's involved in that recruiting process, even for the candidates because

SUMEET: you can also kind of give feedback pinpointed feedback to candidates saying that okay we rejected your resume for this role because of these reasons because right now what you get back and i'm probably guilty of that as well just like a plane that you did not fit the role no feedback no nothing so it it it it it would be beneficial for those candidates as well that they get the right feedback because of that why you didn't get selected, right?

SUMEET: So that like applying the next time they're more careful about it or they'll probably just not apply for a similar role because they know that they'll get rejected, right?

TIM: Yes.

SUMEET: So that just brings in more efficiency overall in the system for everyone.

TIM: Yes, yes, i I hope that is what happens. My fear is that even if companies had a tool which would allow them to give that feedback to candidates saying, here's the main three reasons why your CV was not selected.

TIM: I wonder whether they would not have that turned on, whether they would still have the opaque, sorry, I called a sorry, not sorry email, the rejection email they get.

SUMEET: Yeah.

TIM: my My fear is they might ah just be kind of conservative in giving that feedback. The main fears I hear is around like in some markets, you can be sued,

TIM: for discriminating against a candidate for whatever reason, which is fair enough. And so they tend to be a little bit more sensitive about the one in a million chance that one candidate might miss, misconstrue some feedback and take it negatively.

TIM: But hopefully, as you say, it provides an opportunity for the candidates to get some some kind of feedback, because that would be so valuable, because that is from a candidate's perspective, that is the biggest drop off as well is they make 100 applications to get two interviews, if they suddenly have insights onto the 98, why they haven't

SUMEET: Yeah.

TIM: gotten to that next stage, that would be stupendously valuable, I think for them.

SUMEET: yeah Yeah, absolutely. and And I think it will just like prevent them blindly applying to each and every role that they feel that they'll fit even if they take two of the 10 boxes that are available. right So it did just would help to bring in overall efficiency, save of time for practically everyone involved in this process.

TIM: Yeah. hopefully also, uh, it would Like I feel like at the moment, the kind of narrative in the hiring market is one of increasing distrust and paranoia. Because on one side, every hiring manager would say, Oh, kind of like you said, I'm dealing with all these CVs, a lot of which don't seem to match reality. So there's like a sense that the average candidate is maybe exaggerating or lying even more than normal. Then on the other side of the market, it's the candidates who are applying to so many jobs and getting those callbacks and

TIM: there used to be this thing around all the ATS is filtering me out, which was bullshit, but now suddenly is true. So it's like that the paranoia has come has come true, because now the AI tools are doing automatic filtering. So this is like increasing level of distrust. I wonder whether the route out of that is more transparency, instead of just getting like, a Oh, sorry, you didn't get to the next round, which could then fuel further paranoia. It's like, no, no, no, your CV got rejected for XYZ reasons. Here's exactly why. Hopefully that would then start to build a bit more trust if people understand exactly what's happening in the system, why their CV is not being selected. but What do you think?

SUMEET: Yeah, exactly. I think that mistrust, especially for candidates is there and I've been like on both sides of the table is because like they do not get the right feedback, right? If they are able to get that right feedback, I think like you mentioned there'd be like one off case who made kind of see it like negatively, but it'll still like benefit 99% of the other candidates, right? It'll just give them the right insight as to what they should work on. So, for example, if you reject a candidate's kind of resume, saying that you do not have enough experience when it comes to predictive modeling, for example, or machine learning. So the candidate knows that, you know, probably let me go back, take a course in machine learning, let me enhance my skills in machine learning and predictive modeling. and

SUMEET: then I can apply for similar roles or probably just ah come back a year down the line and apply to the same company and then you have like probably higher chance of getting an interview call. So these things, bringing in more transparency into the overall process, I think it will definitely, definitely help everyone.

TIM: Yeah, absolutely. What about on the side of the table of a candidate in your career? Do you have any particularly strong memories of trying to get a role like any any kind of unusual hiring process or any particularly good hiring process, anything anything struck out or struck out to you as especially memorable?

SUMEET: Yeah, so it was back in 2018. I came, I was working in a company which was like very well structured in terms of data, all clean data, et cetera, et cetera. And then it was back in India and then it was like the startup ecosystem was booming in India. And I was really intrigued and wanted to pivot my ah kind of profile into startups and working within the analytics and growth space within startups. And then I kind of applied for one of the roles and the

SUMEET: The feedback that I had gotten at that point of time was that the startup world and the startup data is like messy. like It's just like it's just done there in data lakes and data warehouses and then you have to do a lot of processing till the time of the kind of ah reach to the end point. And then 90% of the stuff is that's what you would be doing. But I was like I'm faced by that. and I had kind of a technical skills-based assessment that I gave at that point of time for that organization. And even though like I had zero experience for working in the startups, just because ah I had kind of the technical expertise and that was tested well through that system, I kind of went ahead. I interviewed with the hiring manager, I interviewed with the

SUMEET: with the founders and they found me like a good cultural fit as well beyond the technical fit and ah I think I got the offer I accepted it and ah it was the most kind of fulfilling and enriching working experience of my career till date like practically did anything and everything that's how startups work within that space so just because like and had like a hiring or a recruiter seen my CV and like 99% sure that they would have not gone ahead with it just because I didn't have in their eyes a startup experience or the ability to kind of work through like chunks of messy data and stuff.

SUMEET: So and this is what like I have tried to kind of imbibe in myself as a hiring manager that like don't kind of and if if I really see like a candidate there don't just like reject them like at at the very first level itself because usually the first level is just technical skills so if the technical skills are there ah After that, it's just about like having the right attitude to even learn more technical skills, for example, which are being leveraged in your organization and stuff, ah the cultural fitment. But I think that really is something that I follow and have stuck to closely over the subsequent years. So yeah.

TIM: Yeah, I feel like sometimes we don't give people enough credit for their adaptability. Like we just say, oh, well, no start-up experience equals, oh, they don't have a growth mindset or they can't deal with the chaos or whatever, ah but people can adapt surely.

TIM: Like I think of even something like the start of COVID where overnight, almost the whole world went from working in an office to working from home. people manage that pretty well, like on on average.

SUMEET: Exactly.

TIM: so yeah, I think we have a, almost like a built in assumption. Ah, well, if they don't have the experience yet, therefore they can't get it. It's just, which is a bit stupid. If you think about it, because everyone has to have their first experience at some point.

SUMEET: Yeah, exactly. And this is where the the subjective biases come into picture and ah definitely with technology those subjective biases can be eliminated.

TIM: speaking of that, uh, so we've talked about. Yeah, the opportunity for AI to do that initial filtering in a more objective, consistent way compared to a, a tired recruiter who might have 10 seconds to spend on a CV and they've just read 300 of them and their dog is barking at them and blah, blah, blah, all that stress AI systems don't feel stressed. Thankfully. Uh, have you got any thoughts around how to generally make hiring a bit fairer and a bit more objective?

SUMEET: oh Yes, I think of just the initial screening of candidates, I think it and it's time to just revolutionize it. ah a recruiter should not be the first person kind of sifting through CVs. I think it needs, at least for technical roles, it it needs to start off with like a technical assessment, like you've applied for a role, here's a technical assessment, irrespective, like everyone should get that ah a technical assessment done, who has applied, and post that whoever is like,

SUMEET: passes through the criteria whatever you have set in that technical assessment and post that your resumes are kind of evaluated by a recruiter to see like other aspects of fitment and stuff like that. I think ah that will considerably reduce the burden on the recruiter. That will considerably enhance the recruitment process for everyone. I know there would be like people saying that you're just taking out the human element from recruiting because you're not even like letting ah or giving someone a chance to explain or kind of showcase their ability just through a technical assessment you're kind of

SUMEET: ah doing away with 90% of the resumes for example but I think that's necessary and important given the times we are moving in ahead and there is so much of dependency in general on the uses of technology in our day-to-day lives and with respect to like practically every ah role that's there across companies.

SUMEET: So that can be curated. I think the only key is that those technical assessments need to be curated based upon those roles specific to a data analyst, specific to like even an admin or an EA kind of a role specific to a head of analytics role, whatever. So those assessments need to be curated for those levels so that there's like a fair evaluation based on what you're doing. I think post that, obviously your behavioral assessments, your interviews with with the hiring managers and stuff, and then making sure that you're giving feedback to candidates who are not successful at whatever stage so that they really know what went wrong and ah that they can leverage that feedback to kind of improve upon their skill sets for future applications.

TIM: Yeah, I think it's just getting really kind of to the basics of it, isn't it? It's like, is our process accurate and fair? Have we given good and useful feedback to candidates?

TIM: Have we automated away tedious bullshit? I think there's such a, yeah such a ah way to go in hiring.

SUMEET: Yeah.

TIM: What about particularly, go ahead, sorry.

SUMEET: and ah Sorry, sorry, I was cutting out. And just one more thing that came to mind, just like kind of ah when recruiters have their eyeballs on CDs, just like ah not having personal details of the candidate on it, their names, their email addresses, whatever stuff, just from like a DEI perspective, because I feel, and like there are reports that they're like inherent biases, we as humans are biased.

SUMEET: And some way or the other, those biases would kind of come into picture, which I at least should not. ah But yeah I think to eliminate those subjective biases, I think that's another step that can be done.

TIM: Yeah, for sure. And there's now been so many studies in so many different countries looking into this that are pretty irrefutable. And it's actually a really easy experiment to run if you get a whole bunch of CVs that are similar to each other, but with just different names and apply on mass to thousands of different jobs and measure the callback rate, that tells you all you need to know.

SUMEET: Yeah.

TIM: And so I've seen this in Australia and the US s in and the United Kingdom.

SUMEET: Right.

TIM: And it's very, very easy to see where the bias is. And so the fact that most companies still start with human CV, CV containing all this stuff that doesn't matter about the person's religion, background age, all this shit is just ah such a flawed system that we have to change as soon as possible. I think if we want to be fair and objective.

SUMEET: Absolutely.

TIM: what about specifically using data? Like, is there any, we've talked about a few different aspects, but is there any, any element of the hiring process where you think, oh, there's like this new metric, we're not currently tracking or like a new data set that we should be gathering in the process that could somehow enhance hiring if you if you thought about that much.

SUMEET: oh I think from a data perspective,

SUMEET: To be honest,

SUMEET: I'm thinking through my head.

TIM: And if nothing springs to mind, that's that's fine.

SUMEET: yeah I think nothing was like coming to my nose of now.

TIM: What about ah if you think back across your career, Do you have anyone you would consider almost like a hiring hero? So someone who perhaps you learned a lot about how to hire well from, or someone who you'd gone through their hiring experience, you thought, Oh, wow, that was a bit different. That was a bit better. Any, anyone who's kind of springing to mind at the moment.

SUMEET: oh i think work with a lot of senior folks. ah I think for me they they've all had very different styles and different approaches. Some folks have kind of focused on technical skills, some folks have ah focus more on their soft skills with the thought process that technical skills can be learned at the end of the day as long as like someone has the right attitude and the willingness to learn. So I would not like

SUMEET: take any names, but like as a team, I think and that's something that's worked well for me. I think technical assessments are important, are very crucial, ah but at the same time, like I think given the speed at which the tools and the technologies are changing today, like we have Gen AI now, I don't know, probably two or three years down the line, we'll have something totally new in terms of technology so it's about adaptability of a candidate to kind of learn new stuff learn new tools learn technologies the willingness to do it that is something which is kind of crucial for me as a hiring manager and yeah so ah paramount is obviously the technical skills and post that their ability to kind of ah

SUMEET: switch through evolving technologies and being comfortable in this evolving environment.

TIM: Yeah. Anything else you'd like to add at the moment? We've covered off everything I thought of, but in anything else you'd like to discuss in the next few minutes?

SUMEET: ah Honestly, no, I think that's it's been like a great conversation. We touched on a lot of points. I think the key takeaway is that There are a lot of subjective ah decisions that are made in in the hiring process. And ah it's probably key to eliminate those subjective decision makings and make stuff as much objective as possible, even if it is like at a cost of probably lesser human interaction within this kind of space. But then I think it's

SUMEET: it'll just make the whole system much, much more efficient. I'm pretty sure that like the hiring times from can go down from three to four weeks to even like less than two weeks, for example. So it just making it more objective and data-driven rather than keeping it based on gut feelings and subjective biases.