How to build an effective technology team with Mekin Maheswari, founder of Udhyam

Mekin Maheshwari Founder and CEO of Udhyam chats with Amit Somani, Managing Partner Prime Venture Partners.

They discuss hiring and building a technology team, lesson learned while heading HR at Flipkart, problem Mekin's startup Udhyam is trying to solve and more.

An engineer by education, Mekin has had entrepreneurial roles throughout his career. Notably, he joined Flipkart in 2009 to head technology. After having built a strong technology foundation and a great team, he moved on to running two new businesses before finally heading HR for his last 2 years at Flipkart.

Listen to the podcast to learn about

0:45 - Mekin’s career journey. Yahoo, Flipkart & now Udhyam
3:10 - What made Mekin join Flipkart in it’s early days?
8:30 - 'Product market fit is not enough, you also need to find engineering fit'
10:55 - Building great teams - Lesson learned, best practices & mistakes
16:40 - Mekin’s transition from heading engineering to heading HR at Flipkart
20:25 - 'Unlike engineering you can’t undo a decision in HR easily'
22:15 - What is Udhyam & what is it trying to accomplish?

Book recommendations by Mekin
Opportunity for Delhi entrepreneurs to interact with school students as part of Entrepreneurship Mindset Curriculum by Udhyam.

Read the full transcript below

Amit Somani 0:22

Welcome to the Prime Venture Partners podcast. We are delighted to have with us, Mekin Maheshwari. He’s been a very successful entrepreneur. He’s had a long and successful stint at Flipkart. Welcome Mekin to the podcast.

Mekin Maheshwari 0:35

Thank you Amit, happy to be here.

Amit Somani 0:37

Mekin why don’t you tell us a little bit about your career journey pre and post Flipkart.

Mekin Maheshwari 0:42

Sure. So I started my career. I did my engineering from Bangalore and post that joined Yahoo when Yahoo Bangalore was about 20 people. Grew with Yahoo Bangalore. Was there for about four years, worked in data long before it was cool to work in data. Build mobile apps on SMS and but overall found Yahoo to be too slow. Quit Yahoo in 2006, to join a startup as the first hire, startup was called Eugenie. I was there for about three years, we were acquired by an American company after two years. And again, I felt that things were moving far too slowly. I quit Eugenie. And then was hoping to do my own startup, looking for a business Co-Founder, “ki technology mujhe aati hai.” I get tech. But if only I could find a business Co-Founder, I could do my own startup. In the interim, I had attempted a couple of conversations with a couple of people and we’d approach some VCs, who shooed us away. But maybe that’s another day. In 2009, I was introduced to Sachin & Binny by Abhishek, Abhishek Goel of Tracxn who had joined Yahoo alongside me back in 2002. And he was amongst the people who was pushing Accel to ride the first check for Flipkart. And having done that, he was saying, “yaar mere ko technology aati toh mai kar leta,” but given that, you know, tech you know how to build teams. This seems like a great opportunity. why don’t you go and meet them. Met them and then we probably spend as an like they say rest is history. Post Flipkart, I took what most people would call a gap year or a year to understand myself a little bit. This time, specifically within education, what role could I play, and then started Udhyam Learning Foundation. Udhyam is been now running about three years, focused on building entrepreneurial mindsets amongst young people and enabling small businesses to succeed.

Amit Somani 2:43

Sounds great! Mekin, I vividly remember us meeting around 2008-09 timeframe as you had just gotten the offer from Flipkart at a cafe coffee day and hundred feet road in Koramangala.

Mekin Maheshwari 2:55


Amit Somani 2:55

Like you said, the rest is history. And I remember some of the trade offs We’re thinking about in terms of making that decision. You also wrote a blog post about it. So can you talk a little bit about what made you join Flipkart?

Mekin Maheshwari 3:07

So I basically was finding ways to convince myself and convince others why I should join Flipkart, I think, at my gut. So the blog post was interestingly called gut feel over logical reasonings. So to me, the decision was fairly clear. It was fairly obvious that I should be joining Flipkart, but to explain it to my dad. And it’s interesting, there is a some short story to it that when I was 16, my dad had placed two offer letters that he had for himself in front of the whole family, which included me my sister and my mother, and was asking for opinions. So since then, I basically had this very strong belief that family has to be on board with career decisions I make, and I was finding it really hard to convince my dad why I should be taking 50% pay cut to join an unheard of company, where’s 75% of my advisors are telling me that ecommerce will not happen in India. So, “data nia tha, lekin” there was a strong gut feel that, hey, the people and the way they were approaching the problem seemed right. And was mostly because of those two. It was the people and the way they were approaching the problem. And a little bit of customer research that I’ve done with myself as a customer and a whole bunch of other people who were their customers at that point.

Amit Somani 4:28

Sounds Great! How about you know in the first few years as the CTO of Flipkart, it is building out the company. Before we get to your, head of HR role, which was itself a unique transition. Can you talk about the early days of tech, and some of the lessons learned for many of our entrepreneur friends who might be listening to this podcast?

Mekin Maheshwari 4:47

Sure. So quick correction, I never took on the CTO title, was VP engineering and then President engineering, but for most of the “janta” there is very little difference between the two I was the technology leader at Flipkart, the early years, when I joined, we were about five engineers, a couple of interns. And we had processes that didn’t exist, or there were little processes. So for example, the website would go down for a couple of hours every night. And that’s the database update for a time where you were able to update prices and so on. And how do you move from there to doing basics, right, like no downtime, that, here we can figure out ways in which you can do simple things, simple database setup. So master slave and then being able to read, being able to switch over and have load balancers. So bringing basic sanity in moving to the right kind of data centers was the initial bit that hey, as in we were a website, if we were a store that was meant to be 24, seven open, there shouldn’t be no reason for us to go down ever. And I think that in some sense was the primary pieces. The next piece was, how do we architect ourselves? And I was very clear. And then over some time, I hired fairly strong leaders who were equally clear that, we needed to be a service oriented architecture, which allowed for different parts of the website different part of the system to move at different paces. And this also resonated fairly deeply with our people philosophy, which I’ll talk about at length where we wanted people who are entrepreneurs themselves, who were leaders who are owners. So search is one person’s baby. And they would look up search and they would write things. So, that move to service oriented architecture from a monolith database design, and link those ground rules in some sense were some of the early parts of tech designing tech foundation link. I think after this, we got to a point where we will clear that flipkart is going to be massive. And it was going to be massive, both in terms of demand and in terms of what we could build in terms of technology. And for that we needed a very, very strong team. So I end up spending a large part of my time in building a great team. And I count that as my largest professional achievement so far, which is the team that got built and the culture that got built in Flipkart tech between 2009 to 2012.

Amit Somani 7:35

Sounds great! So if I look about technology at very early stage startups you end up having two ends of the spectrum in terms of folks who are running technology, whether they’re VP engineering or Co-Founder or CTO or whatever the title may be, which is that either people will over engineer over architect , over design the system for six sigma five, nine reliability and so on, or others who will do it completely on “jugaad” saying, look, let’s get product market fit. First, and if we live to see the day when things are scaling and growing, we can always architect re-architect rebuild, etc. What would you advise some of these sort of earlier stage entrepreneurs in terms of like, what’s a good middle ground if or either side of the spectrum.

Mekin Maheshwari 8:15

In my opinion, it’s okay to start either ways, as long as you know how to get to the other end, if you do not have the capability to engineer and you’re just making do for finding product market fit, you’ll find product market fit, but then the speed at which you need to be able to build needs to be solid, I think of building technology as a constant evolution alongside the product, not disparate from the product. So wherever in areas where you found product market fit, you better also find engineering fit. We don’t necessarily talk about engineering fit or technology fit. At reasonable levels of demand your technology should start becoming scalable, just like your demand has become scalable. Till the time your demand is not scalable, and it is not scaling. It’s okay for you to experiment. It’s okay for you to make do with MVPs both in terms of product as well as tech. But as soon as you have PMF you better have technology fit because then the demand doesn’t wait and we have enough skeletons lying in the internet space of organizations that hit product market fit but did not hit technology fit. Myspace is my favorite example of a company that hit product market fit so well, that they collapsed. Thanks to lack of technology they collapsed. So yeah, I wouldn’t say that you have to start one way or the other.

Amit Somani 9:35

Yeah, I would kind of agree with you except that I think even prematurely over architecting over engineering the product before you got customer but demand side, the value properly sorted can also be a little bit of a challenging thing for an early stage startup because you will over engineer this but you have no idea whether this is going to work for the customers.

Mekin Maheshwari 9:57

I think the challenge there is often about In the process of over engineering the assumptions that you then don’t let go, because as you are designing and architecting in engineering, without having real customer data without having real customer insights, these are only your assumptions that are getting hard coded into the system. And often those assumptions then last fall longer, even in the face of real customer behavior. So that is dangerous if you over engineer too much. Otherwise, tech over engineering, you can solve for. Just like tech under engineering you can solve for. It’s the mindset that you have about, this is the right way of doing it, or this is the wrong way of doing it. That may be a larger challenge.

Amit Somani 10:37

Sounds great! Let’s jump to the interesting comment you made that you consider one of your largest, if not the largest professional accomplishment to be building great teams. Can you talk a little bit about you know what some of the lessons learned were, what some of your best practices were in terms of building teams and things that you perhaps didn’t get right, as you went along the journey?

Mekin Maheshwari 10:57

Yeah, I think first was just to be ambitious. Yes, that you can actually attract great talent and be proud of the mission that you’re working for. And go after the best talent that you can find. So something fairly simple hack that we did was like people who were doing well at Flipkart, we would ask them who are the five best people that you’ve worked with? And then we would go after them. So that’s one piece. I think second is just having a fairly strong hiring process, which was thought through. And “yahaan”, there are some stories that come with you might be interested in because in early Flipkart days between 2009 and 2010. We ended up interviewing close to 25 people from Google and hired zero. And we got a very, very bad name for it, “tum log apne aap ko smajhte kya ho?” types. Like, who are these people, what is this company that interviews Google engineers and doesn’t hire them and when I reflect back now, that for most of our interviews, we had the technical pieces, where a lot of the Google engineers excelled at but we also had non technical bits where for building something like Flipkart we are very clear that people should have iterated post product market fit, people needed to understand engineers needed to understand the impact of their decision on customers, and where and how that might have happened. And for whatever reasons that the 20 odd people that we interviewed, we did not really feel that on those areas, we were finding fitments, I feel that having an opinion on what you want to build, like ideally, having a thought through opinion on what kind of an organization do you want to build and how, and then being able to implement it through processes, even if it stands against traditional wisdom that Google engineers are the best, I feel is an important piece or is an important lesson. So first important lesson is just think about what kind of an organization do you want to build and then design systems and processes especially high In performance, around those core design principles.

Amit Somani 13:03

Got it. So let’s sort of drill down into that a little bit. Because you started so early, they wouldn’t have been that many people that would necessarily have lived and seen the scale past product market fit. Now, of course, we see a lot of Flipkart and Ex-Flipkart mafia and some of the other internet, large companies where people have come out and have learnt at 3,4,5 years of Flipkart experience, but back then you perhaps didn’t have a lot of that. So how did you then kind of build that early founding engineering group of 10,20,50, people?

Mekin Maheshwari 13:35

We were basically looking for foundational things, at a very very core, simple problem solving design and architecture level were people able to understand the limitations of their thought of the specific solutions that they were talking about. We weren’t looking for people who knew a specific technology. So we never hired a Java engineer. And we’re basically looking for people who could solve a problem. And know up to what limitations their solutions work. So know the challenges with their solutions. And ideally, in some cases, be willing to learn other methodologies or other ways in which the solutions could be scaled. I think it was primarily this bit of been able to learn because like you said, most people in India had not necessarily already developed large distributed systems. A lot of large distributed systems got built barring a few internet companies that were there. So I had done some at Yahoo. We’ve seen some other companies do a little bit but there were very few of those. A lot of Flipkart engineers saw scale for the first time in Flipkart. But that was okay as long as they were willing to learn. They had this attitude towards understanding the why of things. This is a good solution. But why is this a good solution and up to what limits Is this a good solution? So our interview conversations, were, if i may leap, they were often Socratic in nature. “aacha ye jo bol rahe ho aap ye kyun bol rahe ho?” just because something is a best practice. So the phrase best practice was actually not looked that kindly within Flipkart. You were expected to understand why is this a best practice. And specifically in technology was the expectation that you would understand that this is a best practice for these situations it works here. And in some of these situations, it starts breaking because when you had people. So I’ll give you an example of if you had to design for failure. That’s not a common practice in very very reliable systems. And especially if you have to design for failure in process, the part that the publisher is saying that the book will be available tomorrow. But you know that that’s not reliable. Now, how do you model that if you don’t figure out ways of modeling it, then you will end up creating customer experience that’s not what we would be proud of. So challenges like these, you needed people to be thinking about both the business problems and the technology challenges alongside. And without them being able to do both of these, you would end up creating poor customer experience. So basically, we look for first principles and good understanding of customers and business. The one huge positive for Flipkart was its engineers were also its customers, so we could place ourselves in the customer’s shoes fairly easily.

Amit Somani 16:19

Sounds great! Very fascinating! Switching gears, and now talking about your head of HR, or kind of people leadership role, how did that transition happen? And more interestingly, what are some of the skills you could take with you from whatever you had done prior, primarily as an engineering leader into the head of HR role? And what are the things that a head of HR role taught you that you never knew before?

Mekin Maheshwari 16:42

Sure. So post my head of technology role for a year I was running a couple of interesting startups within Flipkart, one in the digital media space where we had a music app called Flyte and ebooks, and the other one in the payment space it’s called Payzippy. Fortunately or unfortunately, both of them shut down. So I have enough in terms of failures in my resume. But while these were still about 10 months into it Sachin had reached out saying, “nahi ho raha hai HR” with basically we’ve tried hiring a bunch of people from outside and we were unable to find people who could make simple things work for us. When I took on the HR role, our largest pain was we were hiring about five people a week for our supply chain where we needed to be hiring about 500 people a week. So how do you crank up an engine that’s right now operating at five a week to 500 a week? How do you bring in basic problem solving mindset, break the problem down, figured out where the bottlenecks are, as you new seek a little bit of traditional wisdom, because not a lot of HR companies that hired 500 people a week or those who are doing it, we’re doing it in many different spaces. So even here, being able to apply fresh thoughts to how you may solve the challenge or what may work what may not work was an important piece. And given my own passion and given, I count my largest achievement as having built the Flipkart tech team. This was a no brainer for me personally, that having an opportunity to do that at Flipkart scale was positive, even though the rest of the world including my own teams, saw it as an emotion. It is funny but that’s how HR is seen.

Amit Somani 18:22

One lesson was basically applying first principles, fresh thinking, technology problem solving to HR, were there any others? And then the flip side, what are the things that HR taught you that you had no idea about before?

Mekin Maheshwari 18:34

Oh, absolutely. Right. The one other lesson is that it’s far easier to do HR when you are the leader of it. So doing HR for the technology team as the technology leader is far easier than doing HR technology when you are not the technology leader. Because unfortunately as HR you do not have access to all the levers. So as a very simple example There was a point in time where we switched our onboarding from a simple two day induction process to a six week onboarding process. Now I had researched it, we had designed it really really well, but for a business function that was looking for 50 people to join them in the month of August, suddenly then those 50 people being unavailable for a month and a half is a huge hit. And to plan for that to design around it is much harder. And hence, decision making reduces quite a bit or becomes slower. I think the other thing that’s very different and very, in some sense very challenging with HR is that, with Tech you can test everything and you can roll back almost all your changes. So you can be highly iterative and experimental in nature, which suits my profile quite a bit. That’s how I operate with HR unfortunately you can’t give salary hikes and say,”aacha haan hum A-B test kar ke dekh rahe hain ki kya hota hai?” Or you promote someone and you roll back a promotion aren’t things that you can do. So in some sense, the speed at which you can learn reduces quite a bit because your leeway of experimentation is a lot lower compared to doing the same things in technology. So one of the things that I missed most in my HR role was the speed, that in the technology, I could operate with far higher speed, with HR I couldn’t operate with far higher speed for both reasons that one the implication of the decisions was that they impacted people and real lives. And second, you couldn’t change those decisions easily. You can’t undo a decision in HR easily.

Amit Somani 20:48

Fascinating! Let’s say there are unlikely but let’s say there are HR leaders listening to this podcast like what would you invite them classic traditional HR leaders who’ve done various elements of HR role, supporting business leaders or technology leaders in various companies, later stage, what are the few things you can tell them?

Mekin Maheshwari 21:07

I would think if HR leaders think of themselves as business leaders solving problems using HR tools, they would probably be far more effective than just HR leaders using HR tools, I think the understanding of business problems of core challenges and being able to solve them, versus implementing systems and processes. So why are we doing a performance management system? Should our performance management system be developmental in nature be evaluative in nature? When to give a bonus and meant to give a salary hike, and in a new startup’s case, how do you even think about the ESOPS? So what’s the mental model of capturing value that you are giving away as part of ESOPS. I think being able to design and think about these from a first principles point of view would be very, very helpful for a lot of HR leaders. If I had to say two words, it will be business-first solve business problems, and use HR as the tools to solve those problems.

Amit Somani 22:09

Sounds great. Let’s switch gears and talk about what you’re doing now. But in particular, what is Udhyam trying to accomplish? And what have you learned in the last few years, you’ve been running it in terms of a grassroot level, talent in India, who is going to be the next generation for entrepreneurs and so forth.

Mekin Maheshwari 22:26

So as I was exploring education, I realized that our education system does not learn. Our education system has not learned much over the last 50-70 years. It hasn’t changed much even though the world outside has gone from being a very violent manufacturing led world to being now a services knowledge based world where careers are changing far more rapidly, but our education system continues to do what it was doing at least 100 years back, and one of the largest efforts of Udhyam has been to enable the education system. And as a result, youth be more in touch with the real world, be more prepared for today. And I feel one of the simplest ways of being more prepared for today and tomorrow is to be more entrepreneurial. That phrase, even though for the startup ecosystem, it’s catchy, it’s sexy, it captures a whole lot of what’s happening to the world around us. Just like an entrepreneur’s world. The world today is changing very rapidly. And like an entrepreneur needs to be prepared and needs to be able to drive the change, or at least adapt to the change, so does everyone else. So what we at Udhyam do is to basically enable people to be more entrepreneurial, especially young people. We do that through building curriculum, which we’ve had success in working with state governments, so that government schools start implementing these curriculums and that curriculum is primarily a curriculum of doing. So it’s not a theoretical book based learning about entrepreneurship or learning about the new world but actually doing stuff. So for example, a key element of the curriculum is students actually doing the real entrepreneurial project with real money. So we at Udhyam have experimented with 3000 students where we’ve given them money to try out, do something entrepreneurial in a two to six week period. And we’ve been fascinated by both the fact that they’ve been able to try something entrepreneurial, and return the money but the approaches they have taken. And in many cases, the kind of decisions that they have made. These aren’t typical innovative ideas, but for people for girls who were unable to talk and introduce themselves to go inside buses, trouble people by selling things to them, has been fascinating personal journeys for them. And we’ve been able to get state governments to do things like that. So we’ve been able to implement Udhyam influence curriculum in a couple of states and are in conversations with about six states right now to get the same done in those state government schools too.

Amit Somani 25:07

Fascinating! Mekin. This has been really good talking to you lots of interesting, exciting insights from Udhyam to Flipkart and beyond. We’ll try and wrap up here. Thank you so much for being on the podcast.

Mekin Maheshwari 25:19

Thank you Amit for having me.

If you believe you are building the next big thing, let’s make it happen.