HARSH PRATAP SINGH

Accepting that shiny job offer?

Okay, so finally I have accepted a full-time offer from an early stage startup. During my journey of talking to founders, HRs, giving 50ish interviews, I have realized a lot. Like a lot for a naive out of uni kid! Also, was helping bunch of folks to clear interviews, so have some third-party perspective as well. AI writing resumes, and AI ATS filtering resumes, what a joke! I have received lots of mails of people who have been reading the stuff I write and it helping them. So, I am writing this blog, according to 2026 reality. I hope the upcoming generation of engineering and researchers have it better than me ;(

tbh, the most accomplished person you know may not have relevant actionable career advice for you. they can offer immense wisdom, but you should also seek out successful people in their early/late 20s who’ve been successful in the current job market. you can’t look at someone’s linkedin trajectory and assume trying to replicate it would plan out the same way now. Things are really different now. But i am writing this as I have collected some wisdom (or you can say patterns) talking to 17 senior engineers and researchers (before saying yes to my first new grad offer I was doing my due diligence for a month or so), so I am sharing it here.

Saying this as a somewhat privileged and maybe a bit entitled new grad, a lot of folks, especially my hardworking friends who want to break into this 2026 cursed tech industry or switch for better roles, and juniors who are more often than not smarter than me, hop on the first offer they can get out of desperation and fear (sometimes excitement as well). It takes courage and leverage to negotiate for sure. I feel it's quite similar to dating girls you don't know much, one wrong move and and you are a gone case (now live with that embarassement boyyy), but if things go right (things will reallly go nice after that). A lot of my smart friends, even after doing everything right, either got there offer revoked, or didn't receive full-time conversion offer in 2026. Too many people going for Masters. Ghosting is literally default right now. After final round, or negotiation period, the HR simply ghosts. The onboarding is either too fast (like literally they want to tomorrow) or too slow (basically offer can be easily revoked). Everything is on fire, hiring is immediate, yada yada, and when it comes to giving out actual offers letter, gone! It seems like candidates have an abusive relationship with companies atp. When they join there lovely startup, initially things look fine, but later they are pushed to do 996 for meager equity (not RSUs). Not everyone can sustain that (especially when you are managing your house yourself, have family responsibilities, are senior, etc you get the point).

Who are you?

In this age of AI, you are not a programmer, but a problem-solver, and the high-level problem to solve is how to make the business go brrr. I dont want to reduce things to just one axis, but simplistically, you are either increasing revenues or reducing costs. That's something to keep in mind. I don't know how to drill this in the minds of my junior ICPC addict nerds and research friends, customers dont really cares about esoteric programming, it's just a bloody proxy. You will find some random Physics major girl coding in her emacs in starbuck like a fukcing psycopath and piss your pants. You are most probably not even a good programmer after you agentic addiction. And, nobody is hiring nerds for having fun with them on yachts. Beautiful software is not the goal. Nor working with latest programming language and technologies. There are companies with idiot HR policies where lack of a buzzword means you won’t be selected. You don’t want to work for them, believe me. Sometimes I think, people are not understanding what good engineers are? Confusion everywhere. Any body who understands one language deeply, will ramp upto another pretty quickly. Just because someone has classical ML experience doesnt mean they can do AI Engineering! You are an asset to make money in a business. Somewhere, think of yourself like that. I love quants because of this. There whole work is to use geeky maths and computers to make money. If they make the firm make billions of dollars, well there bonus is enough to FIRE. Anyways, just stay close to center of activity. Be aware of client being pursued, areas company is ramping up its investments in. It helps in the long run. Information and opportunities are not uniformly distributed. If you are at the right place, you can actually see the future in front of you, and then you will be at the right time (that is become lucky). I decided to join onsite because there is a value in proximity that cant really be recorded in a spreadsheet.

Normal pushback, but brooo what about craftsmanship? Well, you should be able to say :

I did end up writing the credit card billing and accounting system, which is a nontrivial thing to get right. Fortunately for me, just before coming to Google I had taken some time to study computer security and cryptography, so I was actually well prepared for that particular task. ...I designed the billing system to be secure against even a dishonest employee with root access (which is not such an easy thing to do). I have no idea if they are still using my system, but if they are then I'd feel pretty confident that my credit card number was not going to get stolen.

Be good at what you do, not to get a job, but to not stay miserable when working. Too much imposter syndrome is just a sign that you need to go back to studying. Anxiety is not always bad.

The job of engineer is solving business problem with code. But you will most probably never hear an executive say - "The biggest issue we have is that we need to ship 100k lines of code". In my personal experience, biggest problems are with humans, not computers. Dont flaunt tokenmaxxing and agents working in parallel. Flaunt revenue, reliability, customer satisfaction, actual business metrics. Its not how hard you have sweat, its how many mountains you have moved.

Who is getting hired in this fucked up market? And how is the experience?

Most jobs (that you actually want to get) are never available publicly (it's mostly useless to apply to job openings in this day and age, dont do LinkedIn open to work shit), just like most top-tier candidates are simply not available publicly to hire. It's really really fucking hard to hire! smart people just have too many options. and a lot of people (and companies) are shady (i am not hinting towards people not being able to explain their own resume, like literally fake info). Even as a normal person, I am not joking, the number of LinkedIn DMs and emails I get on monthly basis is weird. If you are competent (publicly and anyways tech is small, people know people), founders will throw money at you like a fucking gambler.

And dont be lured with just money (if you are a good engineer, you are making money in long run dont sweat it), pay very close attention to culture. CULTURE is the singlemost important thing that decides your experience in the company. You dont want to be that person who keeps on leaving companies in 6 months collecting joining bonuses. I dont recommend taking money over culture. Just becuase the team seems technically smart, the founders are MIT PHDs (and other things that VCs love like Olympiad medalists, and more), does not mean they are good at managing people (its hard). I have talked to founder who were Directors and VPs at almost all big tech (FAANG+other big ones) while interviewing, and unfortunately, I haven't had much nice experience with some of them (some expect 996, some want me to forward them competing offer letters which is not something you can do tbh, some with 90days notice period, and really common one is getting late to meetings and no remorse for that at all). For some reason, I think as new grad people consider you really desperate? I was not. I clearly had no problems clearing interviews and getting offers. Just because someone has a big title, in this capitalistic world, does not mean they have your best interests in mind.

Again a sidenote, the networking word is thrown around for new grads quite a lot. it just means that you meet people who at some point can do things for you (and hopefully vice versa) and making a favorable impression on them. Help people, don't be a selfish jerk (especially when it comes to someone's employment, it's serious shit). If you ever can’t help someone but know someone who can, pass them to the appropriate person with a recommendation. Be a nice guy/gal. Trust and relationships compound faster than money.

Also, you can forget your GPA after your resume shortlisting. Like forget it after your first job tbh. When someone says you to introduce yourself, dont say GPA is the best thing about me. They are not looking at it, until it's some quant or some wannabe-elite firm. I dont mean fuck up your CGPA, but nobody really gives a damn about difference in 8.3/10 and 8.9/10 tbh. Dont over-optimize it.

Its a weird time for a new grad or early engineer. Companies are evaluating whether they are needed or they are better off blowing that money on a bigger token budget. Maybe I am a bit biased here, but I believe in the value proposition of good engineers. If you are a slopmachine, you are not a good engineer. A good engineer :

Needless to say, the bar to hire is significantly higher. In my personal experience, its easier to get into big tech once you get the interviews, than startups. As I am joining startup, I will write about that. The keything is initiative. Founders dont look for people who need a lot of direction (and oversight) to be productive. They want hungry people (not 996) who have their eyes open and look for opportunities themselves to contribute and have an impact.

LUCK and its anti

Try to make yourself more lucky. Have high AGENCY. Just be bad at making excuses, there are no ends to buts in life. Just get what you want out of life. Intensity at right times have exponential value. An underrated aspect of hardwork is that it induces trust before results. Reliability as a human being goes a long way. You can engineer luck sometimes. Everybody at the end has is a salesperson. We are just selling different things. Its the magnitude of achievements that matter more than frequency of achievements, so take those audacious projects.

But i cant neglect that there is bad luck as well. Its when you are doing everything right, but getting dragged out due to factors beyond your control. This is most common in 2026. People complanining about no luck factor are sitting at 0, which is atleast better than bad luck.

After Meta layoofs, zuck apologized, but people pay with their lives. Losing their paycheck, healtcare, visa (lord help US immigrants), years of career momentum, etc. The leaders dont take accountability of their actions. Meta is literally involved in child sexual abuse in India?! No remorse I guess.

There are severe location advantages

The best job markets are generally in US/UK. If you are based out of New Zealand, its hard for you to land impactful roles. Immigration is a horrendous challenge in 2026! For example, if you are in Swiss studying in ETH Zurich, there are strict laws for remote work (given you wanna work at US/UK based firms are there are comparatively less companies in swiss). So, if possible, stay in US/UK/Europe to maximize your opportunity surface area. India also has a really nice job market for SWE at the top of the hierarchy.

Interviewing

When you interview as a new grad, you don't have much leverage, unfortunately. Most of the time, you will be lowballed. Even if you have a lot of internships and very good work experience, it's not really considered very seriously. As a new grad, I have given:

I have literally built god knows what, 10,000 queries per second level distributed systems, agent sandboxing, more in a bloody 45-minute interview. How well you perform in those interviews is simply not the only factor when deciding whether to give you an offer or not. It is also dependent upon other factors which are not controlled by you. There is always somebody who is willing to work for less money than you are. There might be somebody who has one specific experience and one specific thing that the company wants, which you don't have. There could be a thing where the company has a bias towards graduates of a certain university compared to other universities. There could be a plethora of things that could make them reject you for someone else, even if you are really, really good. There is a lot of luck involved, whether you like it or not.

Almost all the hiring managers, the founders, CEOs, and CTOs shout out of their lungs that they want the best talent, that they want the smartest people, the top 1% of engineers and whatnot. To be very frank and very honest, most of these people themselves are not that smart, they are very normal people with a lot of privilege and the right place at the right time effect. I don't know why software engineering has such ego issues. I know some guys in aerospace and some biotech folks who are extremely competent (they are translating signals from brains to help patients speak, truly noble use of AI), but you will never see them speaking such bullshit with such deep confidence. I guess its VC language. I have talked to people who consider themselves humble, which is paradoxical as a truly humble individual simply cannot internalize that they are humble, otherwise humility will become another garden that ego waters, and indeed a beautiful looking garden it would be.

Understand the paperwork

Dont believe in verbal confirmations, sweet talks, we are gonna be a generational company, we are elite and hire from MIT, Oxford, Harvard, IITs only (this is shit practice, dont be too proud over past achievements, there is only the present). I mean I am not joking every US founder repeats the same script, it's the same damn thing repeated over and over again. Understand what you are signing up for. Like genuinely, take time to understand. Its better that you intern at the company to see what there culture is actually like, and then sign full time offer if possible. This is hard and anxiety driven process, but read the shit paperwork that they are giving, and ask howsoever stupid questions that you want to ask from them.

Everything is negotiable

Engineers with high perceived value make more than those with low perceived value. So, network (and write interesting blogs, so that people come into your emails and DMs offering you an interesting roles, not just Amazon HR DMs). People working in high-cost areas typically make more than people in low-cost areas. People who are skilled in negotiation make more than those who are not.

Overconfidence (not uncontrolled ego) is somewhat a virtue if used consciously. You can fall forward with it. Underconfidence will get you nowhere.

Now, practically, some companies want you to send them competing offer letters. They are shit HRs (or founders), and without second thought run in opposite direction, if possible. Some ask for previous salary slips to find out your market value. Ideally, you should not tell them your competing offers. If they are from some famous companies like Coinbase, Google, etc they already know it from your position. But, if people are asking you such questions (and using it against you), you get the culture of the company. Look for better options.

Startups vs BigTech

If someday you think you want to do a startup, then join one!

This fundamentally is an old way of looking at things. Startups are not as bad as people think of them. Yeah, ESOP are paper money, I am not disagreeing there. But there are things beyond money in careers.

These days, there is an AI research+product lab (which I joined), there is a big tech giant which works with all sorts of things, and some other startups which I dont care about. At least in 2026, a startup is more stable and gives more job security compared to big tech in my personal experience. Big Tech was for stability + brand on resume, but given how many layoffs and how much uncertainty have started to emerge in Big Tech, it's no longer the only thing that you should be considering at all. The best place to work these days is an AI research lab or something like that. It is better to work at Google DeepMind or Microsoft Research or some small AI lab than a company that is still working in the old ways and hiring software engineers and has rigid systems. Also, your big tech tag is not gonna help you much apart from resume screening if you didn't work on something that is worthwhile. I personally know people who just took Google's offer and worked with some goddamn Kotlin instead of taking an offer in a startup, which would have given them a very interesting agentic AI experience that is in much more demand right now. Quality of work really matters. You cant just live with a tag can you? Also, at higher levels, getting into big tech is less competitive (does not mean easier) than SDE-1s. Senior people are less willing to go through interview grinds. You wanna do leetcode instead of having good times with your wife?

I am not saying to go work at startups. There are tradeoffs. But there is no shame in doing that if your offer from Oracle got revoked suddenly. Its better to be employed, then take a stupid masters in US hoping that you will do better next time. Maybe you look desperate when you join startups, but indeed startups are act of some desperation.

Personally, I had very competitive offers in India. I actually have taken a pay cut to join Sqwish in UK! I am not even sure whether we will survive beyond 2 years. Thats the difference that I myself will have to make. And I personally think, its an experience worth having. I will grow at the speed of light, will get opportunity to work with awesome people and shape culture of the company, and solve some really fucking hard technical problems. What else to ask for?

Looking for unicorns

And all the best to people hiring FDEs! The overlap between a great engineer and someone you’d confidently put in front of a customer is too small. The tech worlds hottest new role is hardest to hire. There interview process is beyond broken.

https://www.kalzumeus.com/2011/10/28/dont-call-yourself-a-programmer/ https://www.kalzumeus.com/2012/01/23/salary-negotiation/ https://jvns.ca/blog/2013/12/30/questions-im-asking-in-interviews/ https://paulgraham.com/hiring.html