DEV Community: Tina Bu

Hardest Technical Interview Question You Were Asked In Your Life

Tina Bu — Tue, 03 Sep 2019 19:38:03 +0000

I figured it might be interesting to start a thread of the hardest technical interview question people were asked. I have had quite a few interviews in the past few years and it never amazes me how hard some questions are, even if I am looking for data scientist roles and not SDE.

Disclaimer:
Company names or any confidential interview details will and should not be shared.

Anyway, the hardest question I was asked ever was to parse a string of a mathematical expression with +, -, *, /, (, ) and empty spaces, and calculate the result with consideration of order of operations (Parentheses > Multiplication and Division > Addition and Subtraction). For example,

Input：" 6-4 / 2 "
Output：4

Input: "1+4*3/(2-6)"
Output: -2

It's almost the same question as LeetCode 772 (it seems that you need to be subscribed to see the problem description tho).

For the interview, I have around 45 minutes to solve this recursion problem and handle the case where there is a parenthesis I need to temporarily save the current results in a stack and calculate what's inside the parentheses first, and if I see a - or +, there might be a higher order operator on the right that needs to be calculated first. And not surprisingly I couldn't figure it out. I was having the most problems with keeping adding up to a global result variable but also maintaining temporary local calculation results.

What was the hardest question you were asked during technical interviews? Have you ever used BFS, DFS, Dynamic Programming, Simulation, or even LinkedList in your real life? Do you like algorithm interviews? Are you a hiring manager and have thoughts on how to better evaluate candidates? Do you know how to solve the question I failed?

TODO 9/2 - 9/8 & Review on Last Week

Tina Bu — Mon, 02 Sep 2019 22:27:38 +0000

I am feeling more and more confident about my planning skills as this thread keeps building. Once again I finished all of the 3 top priorities I set up for last week 🥳

KeepTruckin Interview ✅
- Pramp prep
- Prepare questions
- Interview on Friday
Resume ✅
- 1 pomo Kuzushiji
- Finalize and send out to Lynne
Noom daily checkin ✅

I came back from Houston mid-week. It was really fun and the people are super nice. Our HQ is also very cool. All the people there are quite nice and chill. Not sure why our center has to be so old with an ex-electrical technician as our head of digital security.

Interview with KeepTruckin went okay. It's a platform engineer role and I told the hiring manager I don't really have relevant experience. He encouraged me to do the interview anyway and I tried my best. I definitely see my lack of knowledge in API and infrastructure design but I don't think I can do better than that. Especially given it's a 5 round back-to-back interview for 6 hours straight without any break.

I applied for a few roles I am interested in from KeyValues and am very excited to see if I will hear back from any of them soon. I also joined a job hunting group and hope to learn and contribute in the community. My H1B RFE documents were received by USCIS too and hopefully it can get approved.

My weight loss plan with Noom is going well too. Much better than I expected actually. I was going to cancel the plan after the free trial but I might consider keeping it now. I talked about it on Twitter but somehow the html link for the tweet doesn't work here 😿

I am dedicating this week to job hunting as well. And my mantra is, ba dum ba dum bam, dream big!

I tried to apply to some big companies but never heard back from them so I kinda gave up. But I don't want to tell myself that I am not capable of getting attractions from all kinds of roles so will definitely apply to a few more again this week and ask for referrals whenever I can. My 3 priorities this week are:

apply to 2+ jobs per day
- research big tech companies for data scientist roles
- incomplete list: LinkedIn, Twitter, Google, Amazon, Uber
- ask for referrals/apply
Kuzushiji
- 2 pomo daily
Dingpy Article
- revise
- submit to publication

Another exciting thing is I am going to a live podcast session on Friday in New York! The podcast is called By The Book and I have been binge listening since I found out about it.

All right friends see you next week!

Cover photo by Sharon McCutcheon on Unsplash

TODO 8/26 - 9/1 & Review on Last Two Weeks

Tina Bu — Tue, 27 Aug 2019 20:10:01 +0000

I missed last week's review post, oops.

Things has been going really well. First and foremost, I finished all 3 most important items on my list from the previous week.

Driver Attrition Project *
- Finish feature engineering
- Try random forest
- Investigate feature importance
KeepTruckin Interview *
- Pramp practice * 2
Update resume *
- driver attrition
- spacing
- Kuzushiji
- send to career coach for review
- send to friends for referral

We have been finishing up a project for a presentation, for which I traveled to Houston the last Friday. I also had my first round Medium interview which I think went well.

Achievements!

I had my braces removed on Aug 13rd 😁 Invisalign in on 22nd
Last swimming class when I swam a lap 🏊🏻‍♀️
Published an article on interview tips on Medium
Resumed algorithm interview practice on Pramp

Happy Memories

The Handmaid's Tale season 3 ended. Looking forward to the book sequel!
Tried to play volleyball
Visited Baltimore, the historic ships and the Charles Village
team building at Dave & Busters, our team won!
Houston has 85°C, authentic Korean & Japanese & Mexican restaurants!
Visited my Natural History Museum and bought dinosaur shirt

This Week

This week I will focus on enjoying Houston with friends while we are here. Friday I have a whole day virtual onsite interview which should be fun. I also started a weight loss plan with Noom and it seems really cool. Exciting week!

KeepTruckin Interview
- Pramp prep
- Prepare questions
- Interview on Friday
Resume
- 1 pomo Kuzushiji
- Finalize and send out to Lynne
Noom daily checkin

Cover photo by Renáta-Adrienn on Unsplash

TODO 8/12 - 8/18 & Review on Last Week

Tina Bu — Thu, 15 Aug 2019 17:48:03 +0000

Last week has been quite hectic as work really kicked in. But I am very happy about the work I did. I was able to finish all the 3 top priorities I set up 👏. I also had a very productive weekend and went to the gym twice.

Driver Attrition Project* ✅

read legacy code

understand what data we have and what has been done

have a baseline model

Recommender System

Batch Script testing

timezone change to EST

Resume* ✅

update (Monday)

review session (Tuesday)

Medium Interview* ✅

prep self-introduction

prep questions for the hiring manager

Side Project - 2 Pomodoro each (weekend)

hackathon

Kuzushiji

SearchAds

DingPy

Moving on this week, I would like to finish the Driver Attrition Project as much as I can. Also, have an interview on Friday so definitely need to prepare for that. I will visit Baltimore with a friend on the weekend and unplug a little bit from work. My keyword this week is 'productivity'😬.

Driver Attrition Project *
- Finish feature engineering
- Try random forest
- Investigate feature importance
KeepTruckin Interview *
- Pramp practice * 2
Update resume *
- driver attrition
- spacing
- Kuzushiji
- send to career coach for review
- send to friends for referral

Cover photo by Renáta-Adrienn on Unsplash

TODO 8/5 - 8/11 & Review on Last Week

Tina Bu — Mon, 05 Aug 2019 20:42:48 +0000

Some Journalling

Last week was quite productive. I moved to a new apartment with a shorter lease term, getting ready to relocate for my new job. I have decluttered some stuff and have been enjoying the zen a lot.

I also did my second swimming class and now I am able to swim to the middle of the pool. This is a big step (or should we say stroke) for me as swimming has always been one of my biggest insecurities. Now I am in the US so I can just say I have hydrophobia and everyone applaud my courage. I also did 2 yoga classes.

Work-wise, the script to schedule a batch machine learning job on Azure is done and ready for testing. The next step is to store the credentials in KeyVault instead of a config file for better security. After this, my recommender system is officially in production and will be able to auto-refresh itself every day. Hopefully we can get some feedback soon for further model improvement.

I also led another Python study group in my team, where we learned some string manipulation and basic regex. (Check out Python Morsels if you want to level up your Python skills.)

For the hackathon, our team decided on the challenge to create a "reverse mentoring" app which allows more senior folks to also benefit besides only providing help. We are still in the brainstorming phase on how this will work out and it has been quite fun. Have you mentored people and will you be interested in being "reverse mentored"?

BUT, my biggest achievement last week is, I spoke with a mobile engineer at Medium and the next day I received an interview invitation from them! I feel I made a good decision rejecting a company 2 weeks ago because I clearly felt my excitement when I heard back from Medium that I didn't have for that company. I JUST LOVE MEDIUM! (What's a little sad is that Medium is blocked in China and my family won't be able to access it. I think I definitely need to set up a VPN for them next time I go back.)

TODO Review and Plan

Reviewing my checklist from last week, I realized I didn't include any work-related tasks or gym time and didn't allocate enough time for moving so it turned out a little too ambitious. I also realize that I have been consistently procrastinating on a few of the tasks either because I felt it's too challenging or I am not sure what I can get out of it.

TODO from last week

revise resume

read IT Consulting project report: DONE

SearchAds project focus on machine learning piece: DIDN'T DO

add placeholder for hackathon project: DONE

send to friends/mentors for review: WILL DO TODAY

practice summary of qualification for 5 job postings: DIDN'T DO

gender equality hackathon

data source research: WILL DO TODAY

brainstorming ideas -> decide on a topic: DONE

Kaggle competition on Kuzushiji: DIDN'T DO

settle into new apartment: DONE

I will be speaking with my career coach this Tuesday so I hope it will motivate me to work more on my resume, summary of qualifications, and side projects. And I will definitely try to be more mindful and realistic when planning for this week. I also chose 3 top priorities*.

Driver Attrition Project*
- read legacy code
- understand what data we have and what has been done
- have a baseline model
Recommender System
- Batch Script testing
- timezone change to EST
Resume*
- update (Monday)
- review session (Tuesday)
Medium Interview*
- prep self-introduction
- prep questions for the hiring manager
Side Project - 2 Pomodoro each (weekend)
- hackathon
- Kuzushiji
- SearchAds
- DingPy

Mantra

As for my mantra this week, I chose "Vulnerable". And I will embrace my vulnerabilities by asking people to repeat what they said if I didn't hear it clearly and correct people when they misheard me.

Historically this has always been challenging for me. It felt like either I am being a bad listener for not getting their point or being too confrontational to a person who is willing to listen and is trying to understand me. Either way, it felt socially awkward.

Good luck to me for my vulnerability practice and Medium interview!

Cover Photo by Jairo Alzate on Unsplash

TODO 7/29 - 8/4

Tina Bu — Mon, 29 Jul 2019 15:26:51 +0000

My mantra for this week is "Nirvāṇa". (I know, I am a little dramatic 😅)

I don't mean that I have endured great pain or complete destruction. But I did experience a lot of self-doubt since last year. I started my first job and soon realized this is not where I want to be. I blamed myself for not asking some key questions that I should (How do you version control your code? Do you do code review? Do people hang out after work? Is there a dress code? Is there lunch? What's the attrition rate? What's there to do in Delaware?) and ignoring several red flags during the interviews (one of the interviews said "PhDs are not for girls, you should get an MBA" and ended up being the person I report to on a day to day basis.) I compared myself with my fellow CMU classmates who landed a job in Google and Facebook and felt like a failure. I spent thousands of dollars on a big data engineer bootcamp only to realize I don't really want to be a big data engineer I am just unhappy with where I am. I failed to change my job beginning of 2019 so I had to wait almost a whole year till October to try again because of my visa.

It's always been challenging for me to be kind to myself and accept that the old Tina didn't have all the information I do now and she has made the best decision she possibly could.

BUT! I feel much better now after working with my career coach and connecting with a few mentors. I learned about what I want and don't want in my next job. And I felt I am in a much more confident and oriented place to say no to opportunities and focus my energy on a few dream companies. I actually just rejected a pretty good offer with crystal clarity that it is not a perfect match.

I feel I am quite mentally prepared for new challenges and I would like to work on a few things that get me back on track with active job hunting this week:

revise resume
- read IT Consulting project report
- SearchAds project focus on machine learning piece
- add placeholder for hackathon project
- send to friends/mentors for review
practice summary of qualification for 5 job postings
- data scientist
- machine learning engineer
- NLP
- deep learning
- schedule review call with career coach
gender equality hackathon
- data source research
- brainstorming ideas -> decide on a topic
Kaggle competition on Kuzushiji
- data exploration
- call calligraphy teacher
settle into new apartment
- morning routine
- sell stuff, declutter

Thank you @ryan_furrer for motivating me to start my own weekly todo post. I will report back next week on my progress. What's your mantra this week?

AWS Certified Solutions Architect Associate | Exam Tips 2019

Tina Bu — Wed, 24 Jul 2019 22:22:50 +0000

I just passed my AWS Certified Solutions Architect Associate exam 👏🏻👏🏻👏🏻 and I want to share my experience and exam preparation tips with you. Currently, the platform has more than 140 different services and new services are launching every day. It can be very overwhelming trying to understand the whole ecosystem. If you haven’t got a lot of opportunities to use AWS at work, taking the solutions architect associate exam is actually not a bad place to get yourself acquainted with the platform.

In this post, I am going to briefly introduce the different certification exams AWS offers, then talk about resources that helped me to pass my exam, which hopefully will help you pass yours. 🌟

My background: I am a data scientist and my hands-on AWS experience is very limited to mostly EC2. I don’t really work with storage and networking solutions. I certainly don’t design critical large scale high performance production systems. Most of the stuff I learned getting ready for the exam is new to me.

AWS Certification Roadmap (as of May 2019)


AWS Certificationss 2019

As of 2019, there are 3 paths to get AWS certified — Architect, Operations and Developer and there are 5 specialty exams — Networking, Big Data, Security, Machine Learning and Alexa. I believe that you don’t need to clear the associate level to take the professional exam now but I have heard that the professional level is significantly harder so it might be a good idea to start with the associate ones. It also seems to be the case that for the 3 associate level exams, SysOps is the hardest, followed by the Solutions Architect and the Developer track is relatively the easiest.


aCloudGuru Machine Learning Specialty Course slides

A lot of people recommend starting with the solutions architect associate exam as your first certificate because it tests on the most important AWS offerings and focuses on “how services work together, to provide the best (and cost-effective) solution for the situation presented.” And that’s exactly where I chose to start my AWS path. Read this post to learn more about which certification exam you should take.

Solutions Architect Associate Exam Overview

Example question 🧐

Your web application front end consists of multiple EC2 instances behind an Elastic Load Balancer. You configured ELB to perform health checks on these EC2 instances. If an instance fails to pass health checks, which statement will be true?
A. The instance is replaced automatically by the ELB.
B. The instance gets terminated automatically by the ELB.
C. The ELB stops sending traffic to the instance that failed its health check.
D. The instance gets quarantined by the ELB for root cause analysis

The above question is taken from the official AWS sample exam questions. In your real exam, the question is either going to be “given the requirement and constraint, how will you design or improve this system” or “given a system, explaining why certain behavior/bug/bottleneck is happening”. The most important thing is to understand the requirement and constraint in regards to performance, scalability, availability, cost, and security.

There will be 65 questions and you have 130 minutes. All of them are either single or multiple-choice questions. ~65% would generally be a passing score.

My take is that if you have some relative experience with AWS and basic knowledge about networking and database, passing the exam in 1 to 2 months could be a realistic plan (that’s how long it took me). The key is to focus on what’s important — understanding different services, features, and tradeoffs. AWS is not going to test you on which button to click on to set up a Lambda function or how much your reserved EC2 instance is going to cost. But you need to understand what storage solution would be the best choice given that you are storing image data and won’t require frequent retrieval but when you need the data it has to be available immediately.

Solutions Architect Associate Exam Topics

I think the core services you need to understand very well include:

EC2 (AZ, auto-scaling, load balancer, security group)
VPC (subnet, NAT gateway, BastionHost, Network ACL)
S3 (different S3 classes, encryption, versioning, cross-region replication, lifecycle)
RDS (multi-AZ, cross-region replication)

Services that you need to understand their basic functionalities include:

Storage: Glacier, DynamoDB, Storage gateway, Aurora, ElastiCache, Redshift, EFS
Compute: Lambda, ECS, Elastic Beanstalk
Networking: Route53, API Gateway, CloudFront
Management & Monitoring: IAM, CloudWatch, CloudFormation, KMS
Analytics: Kinesis, Athena
Applications: SQS, SNS

I understand it’s a lot of especially with the random 🔥 names AWS people gave their products. What is Snowball ☃️? What is Kinesis 🔬? There is a service called Athena 🧚‍♀️? Why do some services start with Amazon and some start with AWS? Amazon SageMaker but AWS RoboMaker, Amazon Redshift but AWS Auto Scaling? Why completely different things have very similar names? NAT Gateway, Storage Gateway, Internet Gateway, and API Gateway? Why some services have abbreviation names some have spaces in their names some have CamelCase names some even have dashes in their names? Ok “E” is “elastic” in EMR, ECS, EKS, ECR, and EFS but what about ElastiCache (how to pronounce these AWS people?) and Elastic Beanstalk? Shouldn’t it be EC and EB instead? Why there are so many services with “Cloud” in their name? Aren’t everything here cloud ☁️ already? Ok EC2 means 2 Cs and S3 is 3 Ss, but what is Cloud9 and what about Route53? I am going to stop complaining now but do read this post for constructive AWS naming feedback and check out AWS in plain English if you agree AWS naming sucks 😬.

Preparation Resources

aCloudGuru course by Ryan Kroonenburg 👨‍🏫: extremely up to date (the instructor takes the exam frequently and updates the content); gives a great overview of the AWS ecosystem and services; very informative lectures and labs. Highly recommend. Must take. The best AWS preparation course on the whole internet. Only $29 a month what are you thinking just sign up already.
Tutorials Dojo Practice Questions ✍️: one of the only shortcomings for the aCloudGuru course is there are not enough practice questions and the Udemy practice questions worked wonder for me. It resembles the real exam very well and has hundreds of practice questions. whizlabs also offers practice questions but I haven’t tried it myself yet.
AWS mock exam: $20 for 20 questions. You can buy it on the same page you register for the real exam.
Qwiklab: even though you are not going to be quizzed on the steps to create a service or to build a VPC from scratch, I find doing some labs very helpful, especially for the core services I listed above (EC2, VPC, S3, and RDS).
Abhishek Shaha’s article on exam tips: great article on exam tips with example questions but it was written in 2016 so some parts may not be exactly up to date.
A GitHub repo with resources to prepare for the exam: open source repos, blogs & blog posts, ebooks, PDF, whitepapers, video courses, free lecture, slides, sample test, and many other resources. There are some re: Invent videos about system design best practices that I find very helpful.

There are a lot more training resources available that I didn’t cover here simply because I haven’t used them. I have heard great reviews about the LinuxAcademy course and labs. A lot of people and AWS themselves included recommend reading the AWS Q&A pages and whitepapers but they are very long and some content may not be directly relevant for the solutions architect associate exam. Let me know in the comments if I missed anything that you found helpful and would like to share!

Logistics Tips

The exam registration page authenticates you with your Amazon shopping account (weird 🛒). If you have been using a nickname then you need to submit your ID to have it changed.


AWS training and certification sign in

The exam is going to happen at a PSI test location. The closest test center for me is one hour away. I would recommend registering for the afternoon exams so you can avoid the morning traffic. 🌵
Take your government-issued ID 🆔 and a credit card 💳. No electronic IDs allowed. MAKE SURE YOUR LICENSE DIDN’T EXPIRE (No followup questions thank you very much).
Use the restroom 🚽 first because you can’t pause the exam once you start.
You will be given a pencil ✏️, a piece of paper 📃, a calculator 🧮 and earplugs 🔕. I don’t know why a calculator is needed because I was not tested on any pricing question.
You will know if you passed or not immediately after you submit your answers 🥶.
Your certificate is valid for 3️⃣ years. You need to retake the exam or take a professional exam.
Optional random tip: Practice using the mouse with your opposite hand so you can hold on to your pencil for taking notes and save time switching back and forth.

Fun note: There is a Chinese proverb called “左右開弓” 🏹 which means a person who can “draw a bow with both their left and right hand”. It basically means they are ambidextrous and super competitive and smart.

What Now?

Create an account to schedule your real and mock exams: AWS Certification & training
Finish the aCloudGuru course or a course of your choice
Do some labs (If you can create a VPC with Public and Private Subnets (NAT), a highly available web service with ELB and auto-scaling, a blog site with CloudFront as CND and Route53 as DNS then you are very much good to go.)
Take the AWS mock exam 2 weeks before your exam to assess your knowledge and test-taking skills
Practice mock exam questions, a lot of them
Pass your exam and tell me 🙋🏻 how awesome you are
Use AWS in your real life

It’s not a trivial task to pass the solutions architect exam but passing the exam is just the first step. Especially given that the solutions architect associate exam focuses very heavily on concepts and design choices versus implementation details, it is not guaranteed that after you clear the exam you become a great cloud architect. Getting certified is not going to help you quickly debug what’s wrong in your system. It’s not even going to make you remember to shut down services after you are done. There can be a lot of strategies to help you pass an exam but to truly know how to use AWS, use it for your work or side projects is the best way to learn.