I’d like to think I’ve come a long way from where I was a couple years ago. The past few months have reinforced why I chose Computer Science as my major in college. Keeping up with new stuff coming up, trying to learn it, observing tech trends — all these things have sparked the most joy in me the past few months than they have before. Most of these changes happened within the last 2 years, which is why I figured I should think about where I was and where I’m at now.
2 years ago, at amazon
I distinctly remember calling my dad and talking to him about how I hated my job. I was getting no satisfaction from work. Every day, I woke up to a feeling of dread which, looking back, was obviously terrible. On-call rotations were bad and I’d frequently be woken up at night, resulting in migraines later in the day. My manager promised a promotion for 2 cycles, for it to never come up. I was losing motivation fast.
My dad, in very asian fashion, basically told me to suck it up. I was making a lot of money, why did anything else matter? This call was when I realized I did not want to be like my dad. Chasing money was not my passion, nor would it ever be. If I had enough to get by, I could be happy. For the first 20ish years of my life, I was convinced that it was my passion because that’s all that I was taught. Looking back, I think I was very wrong. I am a completely different person from my parents, and that’s okay.
I started trying to prep for interviews, but I was frequently too tired to do so after work. I had stopped going to the gym, skipped breakfast often, and just played games after work because I would want to relax. Maybe it’s part of my personality, but changing things can be really really hard.
Getting laid off
I woke up to an email April 2023 titled “Important information about your role.” Read through it quickly, and found out I had been laid off. I messaged my manager, had a brief chat (she also did not see it coming), and had a meeting with the GM of our organization scheduled on my calendar. This was the first and last time I ever talked to this guy. He was chill, I guess.
At first (and only really for the first 2 hours or so), I was really confused and worried. What would I do about rent (even though I had saved up enough)? Would I be able to find a job (even though I had built up a decent resume)? Was this a mistake (even though, in retrospect, this was a really good thing to happen to me)?
After the shock wore off, I unironically could not stop smiling. I called my parents, let them know I had been laid off. My mom was crying, and I was laughing, partially because I didn’t know how to react to that (why was she crying???) and partially because I knew this is what I had wanted all along. I had wanted to leave and find a job, and I finally had the time to properly prepare and just focus on things that weren’t work. I think I gave myself a couple of days to relax where I walked around the city, took pictures, met up with people, etc. It was really nice to decompress. This is when I started to remember something I first realized in college — when things get tough/busy/etc. sometimes, the best thing to do is to step back, not think about it, and relax. In school, when I fixated on grades, I found myself doing poorly. When I stopped caring about the grade (which was just stressing me out) and started focusing on the learning, I had way more fun, and as a side effect, my grades got better anyways.
Finding the next thing
I started to set up a schedule for myself to interview prep. I told myself I wouldn’t take a long break / relax too long until I had a job ready to go. The total amount I’d be getting lasted me around 4 months — 2 months I was still technically on payroll, but was locked out of my laptop, 1 month was part of the severance payment, and I had 30 days of PTO to be paid out to me.
My plan was to start leetcoding, but I needed to be scheduled with it. In the past I’d been on/off solving problems, but I was pretty bad at keeping up with it because work was not very predictable. I found what worked for me was to force myself to go to a completely different place from my house. I started to head into La Boulangerie in Hayes Valley every week day and spend at least a few hours doing interview prep there. The 30 minute commute helped me separate my work and chill time, which was crucial to making this work. I don’t have a lot of self control, so doing everything at home meant that I’d fall back to just playing video games or being highly unproductive.
A lot of people hate leetcoding, but I think it’s because they view it as a chore. I agree that it’s not a holistic way to test a candidate, but I don’t think it should be viewed as a pain. The way I see it is as a daily brainteaser. I used to read comics every day in the newspaper, and similarly, leetcode offers one problem every day as the “highlighted” problem for that day. There’s a calendar that keeps track of the ones you’ve solved. To this day, I try my best to solve these as I remember to. Again, to me, it’s not a chore, but something I look forward to. A problem that I may have to think about for a while and even then, not understand it, or a problem that’s simple to me, to the point where I feel better about my ability to code and recognize patterns.
The hard work paid off, and eventually I accepted an offer at Kharon. I didn’t really know too much about the company or the space before, but I knew they worked with graph databases, which I thought was cool and worth learning. The interviewers were really friendly, and even though none of the questions were leetcode-related, the prep I did helped out with the problem solving aspects. I was told my role would be a mix of data engineering and software engineering, with high impact since the team dealt with customer-facing products.
Starting the next thing
Technically, the offer I signed was significantly less in total compensation than my job at Amazon. This mostly has to do with RSUs at Amazon versus an unknown amount for the equity I get at Kharon, but regardless, I can say that I have no regrets regarding the pay. If I had the choice to stay at my last team at Amazon and earn more, I would not choose that option. Even though I did learn a lot there, at the end of the day, it was my first job out of college, and let’s face it, I would have learned a lot anywhere. The scale of things at Amazon is really cool, but at the end of the day it didn’t make up for everything else.
Kharon is wayyyy scrappier than Amazon. We use a lot of open source technologies, and when I started working, I realized just how little I actually know. It also took some time to adjust to the new environment and people — everyone was much nicer than people at Amazon. Surprisingly, they actually seemed happy about working here.
This is where I started learning about a bunch of things that interested me, CICD in the real world, using and contributing to open source, Docker, Kubernetes, etc. At Amazon, almost everything was closed-source and specific to the company. For example, there’s a whole internal build system that more or less abstracts that whole process away from the developers (it was probably created before Docker even existed tbh, not sure).
I started working in Python again, which is cool. It’s always been a really enjoyable language to me, and even when I was mainly using Java at work, I’d always answer leetcode problems in Python. I think it was important that I found a language that I was “fluent” in, because it made learning new frameworks or libraries much easier. I started picking up Pandas and Polars, which I learned a bit about in college, but never got a chance to use at work.
This Blog
At some point, I decided to set up this blog. It was the first big thing I did that was public on my github profile. I remember using github pages a really long time ago and thought it would be cool to make a blog site my new home page. I talk a bit about it in Setting up this blog, but at a high level, I wanted to make it really easy for myself to publish to this site whenever I felt like it. I wanted the freedom to do it from my phone or laptop, and most importantly I wanted to make it free for me to maintain. If it’s free for me to keep up, it should be free for anyone that wants to look at it (something I try to hold up with anything I make).
Most of it was just setting up Quartz, the tool I use to convert markdown to a blog site, but it was really fun! It was really cool navigating to my homepage and seeing the first HelloWorld note I made.
I want to say that this sparked something in me. I think it was the realization that the stuff I do at work, loosely, is really powerful and can help me in my every day life. I started to think about how I could use software engineering to just make my life better. The realization here is that I think I’m similar enough to a bunch of people in the sense that if I make something to make my life better, there’s a high likelihood it can make someone else’s life better too.
Other Projects
After I was just posting on the blog for a while, at one point, I just felt like I didn’t know enough about AI inference. I wanted to try using it, but until then I hadn’t really done much with it. I started using Cursor and Claude, and I was thinking of random things I could make. I ended up with a SF street cleaning reminder website, because there were no alternatives I could use, and others had expressed interest in it.
Honestly, I didn’t really give it much thought. To me, it was just a test in how well AI could make what I wanted versus how well I could make what I had envisioned. Maybe this is more telling about my communication skills, but the AI version was not exactly the best. I detail it more here: Sf Street Cleaning.
I had a lot of fun with that at first, and since I wasn’t really getting what I wanted out of the AI output, I decided to make it myself too. I ended up with a dope site that I still use pretty regularly. I shared it around, and a lot of people seemed to really like the concept (I’m not sure if they use it, but my cloudflare metrics do show a good amount of visitors to the site).
At some point, I also got the domain kaushalpartani.com, which was cool. That’s the kind of stuff you don’t really get to do at work, but once you know how to do it, there’s a lot of value to it. I was also recalling random bits of info from CS 168 at Berkeley, something I enjoyed.
I started doing more to put together the technologies I use at work and the random projects I had in mind — a lot of data transformation and work with open datasets. If I have random thoughts and I can answer them with data, I just try to do it.
In Summary
All of this kinda culminated in me applying to the Recurse Center and hoping to learn more to keep doing what I’m already doing. I don’t really know where I’ll go after that, or how long this phase in my life is going to last, but I really hope it doesn’t fade any time soon. It’s too fun to just be learning and applying random stuff.
If I had to sum up random points to provide as advice it’d probably be these:
- You don’t have to fit into a mold that you don’t want to. Don’t try to be someone you’re not. Do what makes you happy because if you do, you’ll do it best.
- It’s okay to take breaks — days long, weeks long, months long. As long as you promise yourself you’ll balance them out, it’s completely fine. I still struggle with this one because I guilt myself a lot, but eventually I want to get better at this too.
- Find what you’re good at and keep doing it. You’ll get better and better over time.
- Stay motivated by defining why you truly want to do something. Any time you falter, think about the why and see what happens. If you don’t get back to focus, then maybe what you’re doing isn’t as important to you as you thought.
- Keep up with random shit. Use tech twitter, youtube channels, blogs, whatever. Doing something is better than nothing. This is how you can find things that interest you and get your feet wet.
- You don’t need to do it alone. You may be working alone, but seek guidance, advice, etc. from people you know. It really helps when there’s people looking forward to what you’re going to make.
- Do something with a routine. Even if it’s the leetcode daily problem, it’ll lead to you getting better at whatever it is.