How To Be A Professional Career Coach & Accelerator (Top 1% In The World) - w/ Melissa

The main core difference between entry-level positions and mid-level positions is the component of ML design/ML system design. Can you build a recommendation system for let's say Netflix to do so much extra hard work for that be a jump or a promotion? What was motivating you to do all of this or to go through all of this? Why I switched from becoming an applied scientist then to externally as a MLN engineer in Tik Tok is because I'm Nam Pande. This is the ready set do podcast and in this episode my guest is Umang Chri. Umang is a machine learning engineer at Tik Tok and prior to that was an applied scientist at Amazon. Um's fascinating journey shows that career acceleration isn't just about clocking

acceleration isn't just about clocking years after years, but it's about continuously learning, adapting, and being proactively ready to jump onto the next best opportunity. We cover what it takes to crack ML roles in big tech companies in the US, even without any prior experience, and spend a lot of time discussing the exact blueprint you can follow to make both internal and external jumps to accelerate your career. You obviously need to know Python. You obviously need to know one of the serving languages which can be let's say Golang or it can be Java or C++. Then you obviously need to understand a lot of technologies. Let's say like if you can walk through probably the Tik Tok interview process, I think that would be super super helpful.

would be super super helpful. I also transitioned my role from ML engineer to applied scientist within Amazon. That jump in Amazon comes with a pay hike which is around 20 to 30% increase in your overall salary at the same level. In line with the theme of learning from somebody who's just a few steps ahead, my goal with this episode is to supply you with all the tools and blueprint you need to accelerate your career progression without being stuck in a rut for years. I came to the US without any other full-time experience.

without any other full-time experience. I straight away came after my bachelors. Subscribe on YouTube and any of your favorite podcast apps for weekly episodes and daily clips from those episodes on YouTube and Instagram. And now without any further ado, here's Oman. Oman. Welcome to the only podcast in the world featuring stories of high agency individuals who are just a few steps ahead of us. Umang, welcome. Hey. Hey. What's up? Doing good, man. It feels like only yesterday. you know that I met you at open atlas and except it wasn't yesterday right because it's been more than a month now actually almost exactly a month ago right exactly it's been exact one month uh I think it was August 15th or 16th where I also interviewed you and then we also shot a minor like clip for your YouTube video as well that's true yeah both of which well at

that's true yeah both of which well at least one of them is in the works and I can I can you know share that bit at least but yeah hopefully I'll be I'm excited to see myself as well on definitely your video will also be getting posted like pretty soon. Yeah. Yeah. And obviously we have a bunch of really exciting topics to cover here but just before I start I wanted to call out your amazing Instagram page um abroad in case anyone is interested in checking that out. Um Um here does a lot of really great street style interviews with really the the Indian diaspora, right?

really the the Indian diaspora, right? Would that be fair to say? Would you agree? I would I would say like majority is like all kind of immigrants who are in the US predominantly but yes since I like interview near like tech offices in the Seattle area so I usually encounter a lot of Indians and few Chinese folks as well right so that's like the majority of my um like diaspora but yes obviously I've interviewed like folks from Spain folks from really uh like southern America Europe as well that's really cool yep so I love watching those I'm just like really gravitated towards any sort of impromptu street style style conversations. I think they're just really engaging. So it's amazing that you know you do this and I'm just you know happy to be connected to somebody that does it so well. So with all that said um I do want

well. So with all that said um I do want to jump into the you know meat and bones of our matter here. So before we you know get into the nitty-g gritties of software engineering, machine learning engineering etc. I want to open up with what is kind of a bit of a tradition on this podcast which is what is your hottest take at least that you can you know say publicly about moving to the US for higher studies it can even be around living in the US or being a working professional in the US but what is a controversial opinion that you think you have you know just in terms of this this sector sector mhm I think it's not a controversial opinion but it should be very simple in mind to have that is you shouldn't

mind to have that is you shouldn't depend on consultances. You should do your own research when you have to come to a country like us. Moving countries, moving continents is I think the second biggest decision of your life apart from let's say number one is like getting married to a partner. Yeah. So that is number one. This is number two. You are changing your lives. Moving your life from one continent to another. So this is an extremely big decision of your lifetime especially in your 20s where the time is time passes like so quickly. So depending on consultancies and depending on other people's advice while not doing your own due diligence can be a major red flag and can can actually get back to you once you let's say come to the US and you're not successful. Um so doing that due

successful. Um so doing that due diligence doing that homework preparing very carefully of what you're going to encounter should be the right step and I strongly believe that even if you come in 2025 2026 a lot of opportunities are still there in the US you just have to be very selective be more thoughtful be more aware be more proactive and take the steps accordingly that's super helpful and I would say that's actually at least for a you know big section of my listeners I'm sure that will be pretty controversial but personally I second that um and to add to what you said I also want to share that this is one of those decisions that you can you know hit control Z on technically right but it's an extremely expensive control Z to hit like it's not

expensive control Z to hit like it's not like you can randomly just be like oh never mind we'll just go back right it's not quite that yeah so it makes sense to your point to really dig deep and figure out exactly what you're getting into. Um, instead of relying on what some random stranger told you who is motivated by completely different um, you know, things and your career is probably not high up on that list, right? Exactly. Exactly. I I I do also want to explore what you said around people that are considering to come in 26 now, right? Cuz we're in the latter half of 25. Um, yeah. Is what would you say is that still a good idea to consider the US specifically? would

to consider the US specifically? would you think given everything that's going on and if you can share why you're still bullish um because I I know there's a lot of you know uh hesitance around this subject right now at least. Yes, definitely. So in terms of let's say if I talk about innovation in terms of if I talk about opportunities in general not just in the tech sector but in all the sectors across the board US is the biggest consumer market. It has lot more companies. It has lot more innovation, a lot more transparency. All of these things are better or if not the best uh in the US as compared to a lot of other like let's say third world countries or if there's other other option is let's say China. So which is

option is let's say China. So which is like more closed in a way as in like you have a language problem you have to integrate into the culture which is more harder here. The best part is in the US you already know the language you are already used to a lot of culture by watching it online. So you are kind of already accustomed and you have to only take few steps to get adjusted to the culture here. The key things that you need to have are hardworking capability and consistency. If you can have these two things once you come to the US and you have a like clear goal, I think you should be able to achieve. If you want a

should be able to achieve. If you want a job, you should be able to achieve that. If you have a longerterm vision of starting a business here, you should be able to do that. And if you just want to experience the culture, experience the diversity, you can also experience that. So it's up to you. Be more uh I would say proactive. Okay. Yes, in this semester of my masters, I want to do X.

semester of my masters, I want to do X. In this se semester of my masters, I want to do Y. So planning that from the get-go. Okay. In the first semester, I want to target for my internships and maybe take lighter courses for my uh coursework for the masters. in the second once I have the internship probably in the second semester I would be taking more harder courses that I really like to learn about but potentially can give me a like lower grade or I might not get a great uh grade. So strategically planning all those things because you are driving every single decision since you uh like land foot in the US. So I think anything is achievable still in the US even in 2025 2026. And uh the other thing that I

2025 2026. And uh the other thing that I still would want to mention is even though majority of people think that the current administration is not helping as to say in terms of legal immigration also uh but I think the entire process is still going towards more merit-based immigration. So if you can provide that value, if you are good at your job, if you are hardworking, you will be uh you will be basically given you will be uh how to explain uh you will be celebrated, you will be given more opportunities. So think from that angle um when you are selecting the universities, when you're applying hard uh when you're applying to all these uh universities. universities. Mhm. Yeah. No, that totally attracts. I myself am also I happen to share the bullishness over over the US as well. At

bullishness over over the US as well. At least compared to as you mentioned some of the other options being Europe obviously. I feel like Europe is actually really the only other option that kind of you know makes sense at this point. I don't know you always have like Australia and such but anyway not to digress too far into uh uncharted waters. Um, I think what you just said also kind of segus really well into I'm curious because you've had such an illustrative career in the US. You've worked for two at least really big tech companies. I'm sure you probably have a really cool internship experience as well, etc. as well. But I am curious um can you walk us through your master's journey? What was it like? And maybe if you can uh share with us maybe one or

you can uh share with us maybe one or two really helpful, you know, not I don't want to say advice cuz you know at most advice I find is usually overrated but maybe we'll put it this way. Things that you wished you knew when you started your masters I think would be a good way to put it. Very interesting. I wish I knew for masters. Okay. Um so how I like so I had like pre-planned everything in my head. Okay. this is what I'm going to do. This is what I'm going to achieve in let's say my masters in my first semester in my second semester. A lot of it did follow the plan but some of it didn't follow. So what I wish I knew number one was yes

what I wish I knew number one was yes the master's course work is harder than you expect. It's not as easy. So I for comparative science right you yes I did my masters in computer science from UMass Ammer which is in like top 15 top 20s in the US. That's huge. Yeah. Yeah. Yeah. Yeah. So sorry please go. Yeah. Yeah. The first thing was like I was expecting master's coursework to be hard but it was actually very hard. U it it took me like one week more than like five six days to complete one assignment even for like uh not super hard courses also. So I used to struggle a bit in that front. But I didn't like uh take

that front. But I didn't like uh take that on me that okay yes I'm bad at it. I just pivoted to accepting that and also uh prioritizing other things which was getting internship getting more opportunities and I came to the US without any other like experience like full-time experience. I straight away came out on a bachelor's so I knew that I would be uh competing against a lot of other people who have two to three years of experience or the same internships, same full-time job position.

same full-time job position. Yeah. So I was proactive on that front from the beginning. Getting more opportunities, reaching out to more folks, getting very aware of what other people are applying to, who are they getting reachouts from, who are the recruiters, maybe I should also reach out, I should also reach out again. So trying to get all those opportunities, going to networking events, career fairs and everything and making opportunities available for myself and making that happen. actually by end of first year uh by end of first semester by December I had two internship offers one at Dell and second at Amazon um and had even interviews at Google and other companies as well by this point. So, so I instead of feeling bad, instead of struggling more on let's say getting good uh

more on let's say getting good uh grades, I pivoted myself on doubling down on my internship prep intern uh getting more opportunities there and uh so if I give you an example in my first semester in one of the courses I got a CGE but I knew the curriculum but I knew the system so well about the university that I converted it into a pass fail grade so that it did not impact my entire CGPA, right? And I still graduated with a CGPA of 3.95 out of four. Um, so which is pretty good at like this uh level, but I mean it's good at any level. Let's be honest.

at any level. Let's be honest. Yes. Um so being that very strategic accepting okay yes even if you don't do well in particular courses there are some ways that you can figure out how to make it uh work for you like in terms of getting a good CGPA getting past fail courses taking easier courses taking courses from other departments not just your department so being very strategic from the beginning about all those things. Got it. So so that is one thing that I wish I knew. Second was I personally felt that I would be integrating integrating so much uh with the US culture like I knew a lot like I wanted to like experience let's say the fests I wanted

experience let's say the fests I wanted to go to tailgates I wanted to um just integrate so much like enjoy that college life and do the party life of the US but one thing that I didn't know was that co would come um so my second basically I was able to enjoy my college life basically for one semester which was also more focused towards course work part-time job getting internship and the entire next year was co basically basically so I'm assuming you joined in 2019 fall that means right correct correct and I graduated in 2020 December so my entire 2020 was me basically staying at home uh basically just living with my roommates and some of the friends of my batch but I still

of the friends of my batch but I still made whatever ever I could out of it. Like I moved in with American roommates in my second second year. I lived with two uh like four of us were living in like two wed apartment. So I was getting that American experience. All three were like Americans other three folks. Yeah. Yeah. Uh and I was getting that American experience every single day living with them. And it's just because I was curious about their culture like how they live, how they approach life, how they uh what are the sports that they like uh how they uh keep the things inside their house, how they everything just I wanted to I was just curious.

just I wanted to I was just curious. I've lived with Indians in host in my undergrad. I've lived with Indians in the first first year of my masters. But I want to experience that. Yeah. No, definitely. I can relate to the thing you said about code. Tell me about your internship experience. How was that like your first time in corporate? Well, anywhere, right? Not even just corporate America cuz I said you didn't have any work at. So, walk us through maybe where you went and what that experience was like.

that experience was like. Yes. Yeah. You know, how you if how you enjoyed it at all. Mhm. So, I did intern at Amazon as a software development engineer and I was in the checkout team of Amazon. Check out is like let's say when you order a particular item you see the final checkout page and then you press like okay yes order select order and then you the payment process after after that place order button. So I was in that team. My entire experience that I had projected in my resume was of like software engineer but like previously like software projects but they were uh but they were mainly focused on the machine learning side.

machine learning side. Oh cool. Wow. like I had taken a lot of ML courses in my masters and I had done an undergraduate like final year thesis at Singapore in Nanyang Technological University. Uh and I worked on a ML like research project there. So my majority of my application was like resume was ML focused but to my surprise I was aligned or like I was assigned to a team which was web development focused and that in a Java JSP tech stack. uh which I was like damn where did I get myself into?

like damn where did I get myself into? But it was still Amazon. So I was like yay. But then I was like nah. So it was like mixed feelings at the beginning and the feelings were similar like throughout the entire like three three and a half months that I uh uh worked with Amazon there. And the main thing that I got from this internship was yes it was very intensive uh lot of like onboarding lot of like understanding like huge um um like code bases and making changes in those areas and like building an entire like new feature for the checkout platform which was like quick look uh like zooming in and zooming out of image like you do on a product page of a um like product. Uh

a product page of a um like product. Uh so that was cool like working end to end like working very closely with product managers ideulating finalizing feature details actually implementing and seeing the execution online experimenting online whether uh users are liking that feature or not liking that feature. So it was a very end to end like uh project and I was owning that end to end. So a lot of skills that I learned in that area like communication uh development process end to end leveraging uh a lot of like uh new technologies including Java JSP which I wasn't like passionate about but I had to do it because I had to get a full-time offer as well. So had to like

full-time offer as well. So had to like work my ass off uh so to say uh get that offer. But eventually yes I was able to get that offer. But the funny part is when I was about to sign my offer I was like I will leave the team as soon as I join Amazon. Even after signing the offer because I didn't want to like work in the software space or like in the I see. So you want pivot internally essentially. So I wanted to go towards the ML side that is where I was interested in and I wasn't interested in this old tech stack of like doing web development in JSP langu like Java server pages or per language was also there and for some piece of code legacy code and then Java. So I was like I just want to switch as soon as possible want

want to switch as soon as possible want to work in Python want to work in ML projects that are like very cool want to get that industry experience. So, so that was my like stint at Amazon as a intern. But then once I joined full-time, yeah, I just like switched after like 2 3 months, gave a lot of like internal interviews for Emily roles and then eventually was able to land uh some positions and then chose one of them and moved to Austin for uh one of the role. Wow. And was that like was the were those internal interviews like pretty hard? I'm I'm assuming they would be pretty rigorous, right? because obviously they're trying to find out a fit and all of that.

fit and all of that. Yes. So they are rigorous but I have to be very honest that they are not as rigorous as compared to like like an external uh external like external rate. Yeah. Yeah. So it's not that extensive but yes we still like they still followed that entire structure of having a coding round. They were having a like oops design round. They were having a ML fundamentals round. So they had like three to four rounds only after that basically they were extending the offers. offers. And did your job title like change completely or was it uh at least similar or you were just working on more machine learning type things versus software?

learning type things versus software? So in Amazon the tricky part is my title was changed. My title was changed from software engineer to ML engineer. But the tricky part in Amazon is ML engineer title still fits under the software engineer family. So there are like different families in I see Amazon where like applied science is a separate family where there you have the applied scientist position but software engineer family has software engineer and ML engineer data scientist is a separate family data science so unfortunately the pay was same the level was same everything was same just the title but the business title was changed to ML engineer and so if somebody's looking to apply you know from scratch to to Amazon like the Amazon MLE position. Could you maybe

the Amazon MLE position. Could you maybe like I I know it's been a few years right for you but could you maybe share just an overview of the type of skills any tools any requirements in terms of their resume that are a must have when it comes to a role like this? Mhm. So in general like in Amazon the ML engineer position is not actually like a regular like ML engineer position at let's say Tik Tok or at Meta or at Google. It's not like that end to end uh ML position where you're actually also building on uh like working on building the ML models. This position is basically in the for most of the cases again it depends from team to team but in majority of cases the ML engineer

in majority of cases the ML engineer position in Amazon is in the context of supporting applied scientists in the entire process of building models. So for example, you will be uh helping in the data ingestion part like getting the data from different teams and then giving it to the applied scientists and they will take care of the transformation of the data and like running the ML models on it and once let's say they have run the model they have evaluated the offline metrics they have that entire thing ready and then you will come in handy again to like productionize that model and like make it uh online. So, so these are the two components where majority of the software like software engineers/ML engineers in Amazon uh contribute uh in an ML team. Uh so the skills required

an ML team. Uh so the skills required for that right so the skills required are you obviously need to know Python you obviously need to know one of the serving languages which can be let's say like Golang or it can be Java or C++ any of the other languages perfect then you obviously need to understand a lot of uh technologies let's say like AWS stack you need to understand at which component you would need to use an AWS Lambda at which you need to use AWS S3 or AWS uh EC2 instances and so on. So you need to know the AWS tech stack and then understand the ML end to end process.

understand the ML end to end process. Um you don't you're not expected to have a depth in one of the ML areas. So that is not something that you will encounter. But this is something you will encounter in an applied scientist interview. Correct. They will be asking you questions about let's say you have a you have mentioned some projects about transformers or like NLP. So they will be asking you dive deep questions about like uh the attention mechanism of the transformer model or like what is the difference between different portions of transformer model and so on. So so they will be asking you deeper questions but as an Emily you're not expected to know all those but you're not you're expected to have the breadth of knowledge exactly exactly from one end to the other. So this is

from one end to the other. So this is the skill set and again like strong coding, strong system design skills if you're looking at mid-level if not then that's not needed but like strong coding oops uh concepts and then ML fundamentals these are the main skills that you would be required and in today's age the additional thing that you would need to know is the latest LLM uh agent AI workflows or LLM workflows because some of the positions require that today. So that is additional skill set that you need to have.

set that you need to have. Amazing. Yeah. So it's so funny um well not funny but interesting that I actually had a a Microsoft applied scientist on the show and this is really good for me because you just completed that circle right like a a lot of the things he had talked about um he had not talked about stuff like you know how do I get the data that I run my models on right and then you came along and you filled that circle for me where you're like okay this is what this person's job is and which is very interesting cuz I had obviously No idea.

had obviously No idea. That's super helpful. And obviously for any listeners, I'll link that in the description too if you want to check that out. But in your case, I am curious. So, so you're at Amazon. Life is goodish. Would you say like did you did you like your time there? Yes, my team uh my like life at Amazon in Prime Video where I was a machine learning engineer initially was pretty good. my manager my so my my I had a manager who was at senior level and like we had some applied scientist folks in the team some uh ML engineer folks some data scientist folks so it was a very broad like group of people who were working towards like building these internal products uh in the forecasting space and the team was very vibrant like very diverse and very

very diverse and very um like I would not say like uh they were very hardwork working but still like very friendly so to have fun. Yeah. Yes. My culture or my team's culture was pretty good in that front and given that it was a co time so we used to have like quarterly offsites. So they used to come to Austin or I used to go to Seattle and once we all went to Bay Area not bay sorry the Irvine area in the SoCal. So all that was like super fun like and lot of like teammates were at my similar age. So definitely it was like a great experience at Amazon at that time.

experience at Amazon at that time. So then I know you're not at Amazon currently. currently. So what happened? Do you mind taking us through? Yeah. So the story before I even jumped there like I also transitioned my role from ML engineer to applied scientist within. Oh you did? Wow. Okay. Interesting. Interesting. Interesting. So because I was working very closely with applied scientists and I had also contributed in some model development work. So my profile was ready in that way that uh like a lot of applied scientists and other MLES in the team also vouched for my transition. So I was able to become an applied scientist as well. well. Is that a normal pipeline like do lots of MLES end up becoming data scientists or applied scientists at least if they want to? Yes, I think if they can easily

want to? Yes, I think if they can easily become that if they are like closely working with applied scientists and they take ownership of one to two let's say model work streams and they start on acting as applied scientists first then they let's say like next performance cycle or next quarter they should be they like if they get like enough support from other applied scientists in the team and like manager they should be able to transition. Yeah. Yeah. Got it. And I'm assuming that that jump comes with a pay hike right? Yeah. Yeah. So that jump and Amazon comes with a pay hike which is around you can expect like 20 to 30% increase in your like overall pretty sizable or same level at the same level.

or same level at the same level. Wow. That's pretty pretty neat all things considered, right? Yep. Yep. Yeah. So the thing like why I switched from becoming an applied scientist then to externally as a ML engineer in Tik Tok is because I was still at it had been like almost two years at Amazon and I recently became an applied scientist at at an entry level. So and then becoming a mid-level would take me at least one one and a half years more if I was even like working super hard. I see. It wasn't going to be earlier than that.

It wasn't going to be earlier than that. But uh now since I had that ML experience since I had the applied scientist uh title as well and I also was working uh for like last 2 years I could externally become a mid-level engineer uh uh in like or like an applied scientist in um in any other company. So so that was my thinking and I started giving interviews and this was my first time giving full-time interviews for all these companies. Um so it was obviously a trial and error gave a lot of full loops at companies including Door Dash, Google um Amazon not sorry Amazon sorry uh like Tik Tok uh a few others as Atlashian as well and I got offers from Atlashian and uh Atlashian kind of like also cleared the Google loop but uh they had like limited

Google loop but uh they had like limited headcount at that time because they were reducing their hiring so that didn't like work out. They were offering me uh L3 position which is the entry level at Google. So I was going to take that. So then I had a choice between atlashian and Tik Tok and then um I chose Tik Tok because the salary was literally 1.5x that I was getting offered at at last year. year. Interesting. Interesting. So yeah that was my journey of like uh externally switching uh from Amazon to Tik Tok and buy.

Amazon to Tik Tok and buy. Amazing. Yeah. So that's super helpful. Automatically my mind is racing with at least a few questions. I think the first being that so when you're trying to make the switch from an entry- level applied scientist to a level two applied scientist is that the term in the industry perfect so when you're trying to do that I'm sure that process is different from when you're first trying to break into your um like opening level positions right so could you help paint the picture of how it's different and maybe if you can walk through probably the Tik Tok interview process just as a reference I think that would be super super helpful.

would be super super helpful. Yes. So the main core difference between let's say entry-level positions and uh mid-level positions is the component of ML design/ML system design. So ML design or ML system design is let's say they will give you a problem of hey can you build a recommendation system for let's say Netflix. Mhm. So now you have to walk through the entire process of let's say gathering the requirements functional non-functional non-functional and then working on what the data is, what the features you will build, what the models you will uh like use and why you will use those and then what the evaluation metrics will be and then tying it back to let's say the business metrics that will

say the business metrics that will impact this entire uh system or the model. So so this is like one additional interview round that you have to go through. Now some companies wanted you to have that discussion around the system area of the entire ML uh end to- end process and some uh companies wanted to have only the problem solving aspect of the uh ML problem like just diving deep into the model details diving deep into the data details gathering details so that depends from company to company but this is additional component that you have to clear like additional one round that you have to clear as compared to the entry- level position.

to the entry- level position. Interesting. And then when you're trying to, you know, prepare for this. So, you have a full-time job at this point, right? This is all on the side of your desk, so to speak. Um, that sounds like a lot to me to do. So, how were you able to allocate time and efficiently use the limited time you had to not just apply, right? But once you have the interviews, you have to study for that is a separate ball game altogether. altogether. So, how how did you manage to do all of those things? Yes. I think the best part is that uh you have to prioritize. Uh I think let me say that again. Uh the most important thing is like in life

the most important thing is like in life the most important thing in life is like setting priorities. So if this is a very big priority for me that switching is a priority for me then I have to make time for it. Yeah. uh and then which means that probably I will not go out on weekends maybe I will uh not uh explo not spend time with friends uh on weekdays in the evenings I will prepare after my job I will spend some time on preparing and the biggest challenge uh for becoming let's say ML engineer or like clearing these interviews today even today in fact at that time it was even a big more challenge is that there is no one resource out there through which you prepare everything let's say for coding

prepare everything let's say for coding for lead code you can just go on lead code you can prepare or for system design you have like still one to two resources you can prepare but for ML the context uh the scope is so vast you have ML fundamentals you have ML coding you have ML design you have ML system design and there's no one resource out there through which you can prepare everything so it was a trial and error again and again like giving interviews getting some idea okay this can also be ours again giving interviews okay this can also be asked this is why I failed. Um and at Tik Tok I failed my full loop three times before I gave the fourth

three times before I gave the fourth time for a different uh team and then I got through all the interviews. So the best part about Tik Tok at that time was there was no cool off period. So once I was getting rejected I applied again and then I got a reach out from a recruiter again uh and then I gave the interviews again. So first first time I failed in the first interview. Second time I failed in the second interview. Third time I failed in the third interview. That's crazy. First time I hear Yes. Yes. Wow. I can't. And so hold on. So did they know that you've been applying and like what you said last time and where

like what you said last time and where you failed and all that, right? Probably. Yes, probably. But did it not they did not bring it up like proactively to talk to you about that? They did not bring that bring any of that up and maybe they cannot even see I think different teams in Tik Tok cannot see what the performance was in the previous team. I see. I see. So and the recruiters are totally different for all these positions. That's true. So because the company's like very big.

So because the company's like very big. So they don't get to know all those details details and um again like the interviewers for ML are like very uh interview like interviewer dependent as well like team dependent as well. So you don't know like what is getting asked. So the day I used to get rejected I used to apply again again and I did not like and this was I think 6 month journey 6 month of entire like trial and error starting April till November starting I started working in December. December. Interesting. Interesting. So it was a long journey um of like preparing giving interviews keeping everything maintaining my job performance as well. So, yep. So, what do you think ended up being the final click that you needed?

being the final click that you needed? Like, was it just a question of multiple attempts? It sounds like it was and it to me it feels like you were the difference that you did something that got you that offer that fourth time, right? And not just a function of the number of attempts that you were getting. getting. Oh yes, definitely. It was again like my preparation for coding. So, let's say initially when I gave interviews, I wasn't fully prepared. I was still weak in some areas but I kept on interviewing and I kept on preparing. So by the end of my let's say when I was giving that final loop for tech talk I was able to complete the two coding questions that they gave me within 30 minutes and the interviewer was so impressed by my performance like they

impressed by my performance like they expect at least two coding questions in like that 45 minute and they also ask you some project based questions in that time. So I was able to do that in the optimal way testing like complexity wise covering everything time overall like going through the cases test cases and covering everything. So the interview was super impressed and same thing happened in the second round as well because I was keeping on preparing. I was consistent with my prep that eventually led me to perform really well in the interviews. Um and this is again like very important even today because right now if you're not doing the best in the interview uh in 2025 2025 it's most likely that you will still get rejected even if you perform at let's say like or you think like you did great

say like or you think like you did great but still there's a very high chance that you will get rejected in today's times because there's a lot more competition lot more people are uh willing to like become MLE and lot of people have the similar performances performances So you need to be at the top of your game uh especially today. But I was able to clear all those interviews at that time just by constantly preparing and improving from my failures like learning where are my missing pieces. Initially it was coding then it was also ML system design then working on that refining my answers uh in an optimal way to save time in terms of behavior interview so that I don't blabber a lot in my uh interview because Tik Tok interviews are only three rounds

Tik Tok interviews are only three rounds and they have in the first round they will ask you some ML pro like your project questions and some about like mainly focus on the coding part in the second round they will focus more on ML coding and some ML fundamentals. And in the third round they will focus more on ML system design or ML design and alongside they will ask you some behavior questions. So you have to not just focus on one component for the entire 45m minute interview but also cover that let's say in 30 minutes and then have some time for other part of the other component of the interview. So you have to be very time efficient. So all of this like went through the trial and errors, practicing, refining my answers, keeping them succinct, giving more value in whatever I say. So that all led to the entire success in the

all led to the entire success in the end. I think one of the things that I'm really blown away by is your continued continued perseverance through this entire thing. And I say that because I don't know I I I speak to a lot of friends of mine and I think this partly applies for me as well but most people when they are at a job any job right and they've been at that job for like a couple years you know life is good well if if not good life is going I guess and and then it just you fall into almost this kind of this rut right where you're like okay it's it's not the greatest but pays the

it's it's not the greatest but pays the bills I'm living comfortably getting by at work blah blah, what's the point of hustling so much harder? So, I guess really what what I'm getting at is for all of any of my listeners that are kind of on the fence about is it worth, you know, going out of their ways to do so much extra hard work, effort effort for that, you know, be it a a jump or a promotion or whatever the case might be.

promotion or whatever the case might be. What was the thing that was really motivating you to do all of this or to go through all of this? Yeah, I think number one was that I've come to the US uh and I know I can make a lot of my time here which is I know I can get a higher salary. So if I'm being very honest salary was one of the biggest factors uh in my journey as well. well. And then also it was more like okay yes I've been in this team for one and a half years now. I know how things work.

half years now. I know how things work. I want a different challenge. I want to work in a new ML domain. Um so this was also another thing that I was from the beginning like interested in like trying out uh like seeing the working working style of different companies. Um and since the beginning you know I always had this thought in my head that I would be working for Google or a big tech company where I would get free food. It might sound very silly but yes I wanted free food in my love free food. Yeah I was my dream and and I got that and I was able to achieve that through Tik Tok. Now I get free lunch, I get free dinners, I get free snacks. So, so this is good. This is a great perk

so this is good. This is a great perk that I save a lot of my time on a daily basis. basis. Um, and then I like higher salary working on an interesting like ML area. Uh, working at such a big scale. Tik Tok is a very big product, biggest product in terms of social media in the US. Yeah. So working on all cool like projects uh and then obviously getting all these benefits making the most of my time here in the US earning a higher salary in big tech. So all of these were my goals and I feel I was eventually able to achieve them. That's amazing and so inspiring you know. Thank you. Um are you able to at all speak on what you do at Tik Tok? Mhm. Yes. So right now so I recently um

Mhm. Yes. So right now so I recently um so I work in the space of uh so let me say that again. So I work in the space of risk management for uh Tik Tok. So basically let's say if there is a user on Tik Tok platform and they are posting some sort of like bad content which can be lowquality content which can be um like some generated sloped like it can be explicit content can be sexual content so all these kind of contents we detect them using machine learning learning and I I am owning the piece of creator rewards program at this point. So which means that Wow. Yes. So I own that entire project. So basically who gets into that entire uh who will be earning revenue and who will be like moved out of that platform based

be like moved out of that platform based on their content based on their like behavior on the platform. If it is a group behavior and like there are tens of hundreds of accounts doing the same kind of like malicious attack on the platform, posting similar content. So removing them, not helping them earn a lot. Uh so I basically own that piece of uh Tik Tok US and yeah basically I build ML solutions and maintain existing ML solutions to uh yeah reduce the risk on the platform and uh improve our like overall revenue because we give revenue to we give so much uh money to all these creators. Exactly. we can save that money on the like bad users then that's an impact of like millions of dollars

an impact of like millions of dollars that we see on a monthly basis. So yeah, so that is what I own. Definitely. That's so interesting. And yeah, like I feel like that's almost probably at least looking into the future, right? One of the most important areas I would say for Tik Tok. Um, I imagine this has been so helpful and you know really appreciate you taking the time and I wouldn't be surprised if there's a few comments asking for various or the topics that we only could a little bit get into and I I'm sure if you'd be open to it I'd love to have you back to you know look at any of those topics in the future.

topics in the future. Oh definitely. I would really love to like get back and talk with you Nan. I really enjoyed that you did not have some sort of script or like any like prepared questions in general. you just had some ideas in mind and you were going off of about what I was responding. So I really felt like it was a very natural like conversation that I had with you and yeah great job on uh this podcast. Thank you. I try.

this podcast. Thank you. I try. That brings us to the end of that fascinating episode with Umang. I took away so much and honestly I think anybody listening is now better prepared to make those big jumps in their career going ahead. If you would like to support me and if you found value in this episode, the easiest way to do that is by subscribing on YouTube and leaving me up to a fivestar rating on Spotify or any of your favorite podcast apps.

any of your favorite podcast apps. Something that goes a really long way for me is if by word of mouth you share this episode with somebody, a friend, a colleague, whoever it might be, and interact, engage, drop a comment for any guest recommendations, any any suggestions that you want me to take going forward in the next episodes. Catch you all in the next one. New episodes every Wednesday.