Lessons from a tech job search


A 365 days in the past I turned into pondering interviewing but I turned into feeling anxious. I hadn’t interviewed in 4 years. My resume wasn’t as much as this point. I had simplest a obscure concept about what firms were on the market and what the map would be like at every. I delay applying as a result of the phobia of the unknown.

I need I had get right of entry to to a primer on what firms were on the market and the trend to map every. How long the map would exhaust from starting up to attain. Programs on how to situation up for every. I would perhaps well no longer get this form of post then, so I believed I’d write it now.

What I chanced on a piece of exasperating in same posts is the reluctance to title the companies enthusiastic. I’m going to title the companies right here to coloration a certain image. As an instance, in account for so that you can interview at firms that don’t state algorithmic interviews, you might know where to coach.

It’s imaginable I’d get some judgement for the companies I did or did no longer state to, or because I obtained rejected by some company. I’m okay with that.


  • Since right here is based mostly entirely entirely entirely on my abilities, or no longer it’s some distance a sample size of 1.
  • It’s been practically a 365 days since I carried out this path of, so a few of the companies would perhaps delight in modified since then.
  • I’m a backend machine engineer.

Getting willing my Resume

I set up this off because I needed to transform it entirely, in conjunction with 4 years value of labor. And then it grew to turn into a standard case of yak shaving. Sooner than writing the resume I situation out to get the obedient tool to write it. After selecting jsonresume because I cherished how it saved inform and structure separate, I chanced on it would no longer work on my M1 laptop laptop. Then I wasted time getting this working and fidgeting with the structure. At closing I wrote and polished the inform.

In hindsight, executed would delight in been better than most interesting. Also, I would perhaps delight in made it more easy for myself by updating my resume after every efficiency cycle when my achievements were fresh in my mind.

It’s now kept in plaintext + git so updating is discreet.

Making state of

There are a bunch of strategies that you might state, and I ancient all of them.

  • Company.com/careers. Apple, Babylon, Cloudflare, Deliveroo, Monzo, Spotify, TrueLayer. (7)
  • Referrals. Amazon, Palantir. (2)
  • Inside recruiter. Google. (1)
  • Exterior recruiters. Instabase, JPMorgan. (2)

Where I did no longer state

  • Didn’t are attempting to work there. Finance/quant roles, startups.
  • Didn’t must coach on the time. Shopify and Transferwise (now Vivid).
  • Didn’t delight in relevant roles in my city on the time. Microsoft, Netflix, Revolut, Stripe, Uber.
  • Already labored there. Fb (Meta).

A spreadsheet allowed me to trace at a have which processes were unruffled ongoing, what the subsequent action on every turned into and detailed notes on all conversations. My memory is uncomfortable, so this turned into a truly great.

I obtained certain responses from most of the companies I reached out to excluding

  • Apple. Automated response, nothing after that.
  • Spotify. Automated response (“that you might exhaust mark to the playlist we’ll play in the set apart of labor as we learn your resume”). To be shiny, the playlist is 🔥.
  • Google. This one turned into unparalleled because the recruiter had emailed me every 6 months going support 3 years. Unfortunately, on the time I reached out the recruiter turned into switching roles and the team it turned into handed to turned into being reorganised. I did no longer know this though. After about a weeks of silence I asked a chum to refer me. I obtained a call precisely a month after my first electronic mail.

Lesson: or no longer it’s imaginable for applications to trudge thru the cracks. I do no longer know if a human ever observed my application at Apple or Spotify. If they did, I do no longer know why they handed. I did no longer know someone there like I did at Google, so I would perhaps well no longer get out extra.

It is a on condition that some applications will be unsuccessful. We mustn’t let it distract us or get us down. Help focussing on the others.

Companies left – 10.

Screening path of

There might be in most cases a screening path of where the interior recruiter applies some authorized filters. These questions mostly revolve around previous jobs, challenges you delight in confronted for your roles, technologies you delight in labored with. “Why are you switching” is a each day quiz.

A couple of firms employed a ordinary filter.

  • Cloudflare asks you to exhaust an on-line refer to authorized tech questions. I chanced on most of them easy excluding one very particular, obscure quiz about SSDs. I grab whereas you answer it, it looks to be like like you looked up the solutions somewhere.
  • JPMorgan. On-line refer to an intermediate self-discipline. It’s been a whereas since I ancient Java but there turned into a few time to learn about up syntax and stuff.

Cleared the total screening processes excluding Palantir. They did no longer train why and I did no longer state up.

This stage will be a chance to get out extra about the function. I learned that the function at Instabase wasn’t a obedient fit for me and stopped the map.

Companies left – 8.

Preliminary interview

That is where it gets more difficult. I expected most firms would state the dreaded algorithm interview in the initial interview. It grew to turn into out I simplest did one algorithm interview and even that turned into a certain abilities. Here is how they all went.

  1. Amazon. I turned into staring at for an algorithm interview. As a replacement it turned into 15 minutes of questions about my abilities adopted by 30 minutes of coding. This turned into less about solving the self-discipline and extra about the trend to structure the code.
  2. Babylon. Questions about work abilities. Extra of a conversation than an interview.
  3. Cloudflare. Asked me to dwell code a controversy inviting TCP sockets in a language of my quite plenty of. I ancient Glide because I have the win kit is discreet to make state of and has fair correct documentation. It went alright but I made about a authorized errors. Coding an queer self-discipline under time stress is no longer the least bit times easy. Exact ample to accelerate though.
  4. Deliveroo. 15 minutes of tech questions. I obtained to resolve 2 issues out of 5 alternatives. I selected “running programs” and “object oriented programming” and answered the total questions. 45 minutes of algorithm interview. It started with a straightforward self-discipline, then a complex state-up after which a extra complex state-up. I did no longer delight in time to code the third but I explained how I’d map it.
  5. Google. This is capable of pleasure in been an algorithm interview on the opposite hand it did no longer happen. On myth of the mixup I discussed earlier, I turned into executed with initial interviews with the total diverse firms. The interviewer at Google turned into accommodating and equipped to rapid discover me to the closing interviews. I gratefully permitted.
  6. JPMorgan. This turned into a piece of of an outlier. I spoke with 2 engineers who asked very particular questions about Java. Admire “when is a static block flee?” I believed it turned into a piece of unparalleled to match effortlessly googled programming language trivia. They evidently felt it turned into a obedient proxy for machine trend capacity. They later obtained support to me asserting they wished any person with abilities with Java 14 and I had simplest labored with Java 8. 🤷‍♂️
  7. Monzo. Receive home check. Of the total interviews, I enjoyed this essentially the most. It turned into a tiny self contained, valid world self-discipline. They suggested spending no extra than 4 hours on it. But the truth is, I had an even time with it. After I executed coding, I ran the linter to get an A+ on code quality. Then I situation up Github Actions to robotically flee the unit assessments because why no longer. Afterwards I discussed the code with an engineer and explained my map. Gigantic dialogue.
  8. TrueLayer. Receive home check – “write a HTTP endpoint that hits these two public APIs, mix the result and return it”. I equipped to post a each day answer and set apart whatever enhancements they asked for. However the recruiter talked about they’d prefer it to be coded to manufacturing standards and I’d simplest get one submission. I figured I’d be at an advantage spending time making willing for algo and assemble interviews as an quite plenty of because that would perhaps well make stronger my chances with multiple firms. I believed I’d arrive support to this later but I did no longer because I had affords by then. Receive home assessments are dazzling with me but long, starting up ended ones usually are no longer easy if I’m interviewing in diverse places.

Handed all excluding JPMorgan and TrueLayer.

Companies left – 6.


I had tried to schedule most of the initial interviews on the identical time, adopted by the total closing interviews in succession. This mostly labored because I turned into ready to situation up for one thing at a time. I also can learn from errors I made at one whereas doing the quite plenty of

But it absolutely would no longer continuously work. I took my time scheduling the closing interview with Cloudflare. In that time they made a guideline to a ordinary candidate which turned into permitted. And never all firms work for your schedule. They would merely no longer delight in interviewer availability in the slot that is convenient for you.

Ideally you need the total processes to stay on the identical time to set apart negotiation more easy, on the opposite hand or no longer it’s no longer easy to pull it off.

From applying to getting affords, it took me 2.5 months. Whilst you happen to sprint every stage or state to fewer firms you might attain it in as small as 1.5 months. Whilst you happen to coach to Google, I’d request on the least 2.5 months, more than seemingly extra. By myself among the companies I discussed, they’ve a extra stage where you interview with likely teams.

Companies left – 5.

Algorithm Interview

How I willing – In 2017 I solved most of the complications on InterviewBit and chanced on it functional. Pretty than doing the identical complications again I definite to exhaust a have at Leetcode. I did approximately 30 complications over the path of 4 weeks. I believed Leetcode top rate turned into value it because that you might get right of entry to complications by company. I did a few complications tagged Amazon and Google sooner than those interviews. Now not the truth is one of many questions asked by any of the interviewers overlapped with the ones I practiced on the opposite hand it unruffled gave me self belief going into the interviews. I did about a mock interviews.

How I’d suggest others put collectively – Practise on either InterviewBit or Leetcode until you are ready to write the code, debug and cure for all cases contained in the cut-off date and with none hints.

My obedient algorithm spherical goes like this:

  1. Eavesdrop on the self-discipline, compare clarifying questions.
  2. Uncover the answer with complexity.
  3. Code the answer whereas explaining my thought path of.
  4. Step thru the code with a check case. Fix complications if any.
  5. Level out or write the quite plenty of check cases.

I need the interviewer to slouch away thinking – I can write code that works and talk technical ideas well.

At Superior Tech firms, algorithm state helps you certain the screening interview and get a foot in the door. Doing well in the closing algorithm rounds are critical to boot, but whether or no longer you get a guideline and the level of your provide relies upon for your efficiency on the assemble and behaviour rounds. And for firms that skip algorithms, those two rounds are the total lot. Prioritise accordingly.

The three coding interviews every with Amazon and Google went well.

Babylon, Monzo and Deliveroo did no longer compare algorithm questions in the closing interview.

Create Interview

How I willing

  • Designing Data Intensive Applications by Martin Kleppman front to help. This e book goes beyond interviews, it’s some distance a truly great reading for backend engineers. It took me about a months to learn this, on the opposite hand it turned into value it.
  • Design Create Interview by Alex Xu which solves some repeatedly asked assemble questions. I chanced on it functional to learn about a days sooner than the interview.
  • Design Create Primer is yet every other helpful resource but I did no longer state it great.
  • Carried out with about a mock interviews.

How I’d suggest others put collectively – Learn these two books. Prepare a pointers of what you’d rob to camouflage in the interview. Sooner than every interview, mentally flee thru how you’d implement their merchandise. As an instance, Fb and details feed.

My pointers for backend assemble interviews would perhaps attend as a spot to begin. I no longer continuously had time to get to the extra credit bits.

Amazon, Babylon, Deliveroo, Google, Monzo – all of them were spell binding discussions. That is my authorized interview to attain and I enjoyed talking to the total interviewers. The rounds went well though I did no longer have so on the time.

Behavioural Interview

How I willing – I had cleared the Amazon interview as soon as sooner than so I knew the drill. I knew that every of the 6 interviewers would compare me for a time I displayed 2 of the 14 management principles. So I’d no longer sleep discussing on the least 12 of the 14. Ideally I’d be ready to take into myth an event where I had executed these objects on the soar, but like I talked about my memory is uncomfortable. It takes me time to bear in mind things. So I sat down and wrote a obedient memoir from my abilities for every of the 14. When interviewers asked, I told them a relevant memoir. It turned into easy because they were fresh in my mind.

Every company asks these questions, no longer correct kind Amazon. All of them are attempting to know how well you work with others, whereas that you might decide to an concept you disagree with, whereas that you might exhaust possession of a controversy home, bring outcomes, mentor others and so forth.

A couple of of the solutions to these questions did no longer show conceal me in the ultimate gentle. I would perhaps look (or imagined I would perhaps look) the interviewer’s body language replace negatively as soon as I talked about those things. But I figured or no longer it’s better to be correct than set apart stuff up.

How I’d suggest others put collectively – Same. Prepare your anecdotes. Be correct.

Reverse questions

Every interviewer asks whereas you delight in questions for them. Here is a vast listing of questions you might get started with. You might per chance get on the least about a questions which will be crucial to you.

But IMO right here is simplest for you, no longer to provoke the interviewer. In my abilities, asking no questions did no longer delight in an impression on the result. As an instance, by the point I did the 3 coding interviews at Google I turned into mentally exhausted and would perhaps well no longer muster the vitality to match them the rest. I unruffled obtained a guideline. Your mileage would perhaps merely fluctuate.

You can also compare these questions after you procure a guideline by soliciting for the recruiter to set apart you enthusiastic with the supervisor or an engineer from the team you’d be joining. They’re in most cases delighted to oblige.

Inquiring for feedback

I asked about a firms if they would merely part interview feedback with me and some did. Others overlooked the compare or in a well mannered capacity declined (Amazon and Google).

Whilst you happen to dwell in a jurisdiction with an details privateness legislation like GDPR, that you might compare all details they’ve kept about you. They are obliged to give it, though you possibility burning bridges with them. I opted no longer to attain this, but I know this capacity works.

I chanced on that I’m a harsher critic of my interview efficiency than my interviewers. It’s fun to match their feedback with what I wrote about my efficiency after the interview. I simplest centered on the negatives, nitpicking whatever diminutive errors I made. They took a broader have, with the positives getting equal air time.

You might per chance learn one thing interesting from your interview feedback. There might be no longer any damage soliciting for it. Admire the thinker Michael Scott talked about – “you accelerate over 100% of the shots you don’t exhaust”.


Whilst you happen to exhaust mark to correct kind one thing on this post, it wishes to be this – you will delight in to unruffled get advice on how simplest to negotiate. These two links are fair correct to get started with:


I obtained affords from the 5 firms I did closing interviews with – Amazon, Babylon, Deliveroo, Google, Monzo.

It turned into a no longer easy decision but I opted to movement with Monzo. I turned into influenced in phase by the vast interview abilities. The function also aligned well with my previous abilities and I have been a pleased buyer for years.

Monzo is currently hiring!


None of this might per chance be imaginable without the support of my friends. They performed mock interviews and helped me make stronger. They supported and impressed me as soon as I doubted myself. Their well timed advice helped me set apart better choices. Thanks to (in alphabetical account for) Anurag Gupta, Arathi Sundarram, Bochra Gharbaoui, Chandra Sekar, Kushagra Sinha, Minesh Patel, Rajat Khanduja, Sri Raghavan, Varsha Krishnan and Vivek Bagade.

Recruiters are delighted to spend time and effort helping you feel as jubilant as imaginable. A couple of the recruiters right here went above and beyond and I’m very grateful to them.

Lastly, yelp out to the total interviewers. I had a vast abilities as a candidate. Interacting with you has also proven me how I can make stronger as an interviewer. Thanks.

Feedback on Reddit and HN

Read More



Hey! look, i give tutorials to all my users and i help them!