Introduction
I've worked as an intern in a small European company, a software engineer (mostly backend) in big Korean company, and finally an intern researcher in IT startup.
What brought me here?
I can say I have approximately 1/4 bachelors degree in computer science but have never studied computer science "seriously". Currently, I am a graduate student in computer science with B.Sc. in biology.
During my university life, I was not that deep into studying from books. Instead, I prefer to share my opinion, discuss with people, and work in the fields (like actual land fields for farmers).
However, as you know, at some point in your last year in uni, you must decide what you will make money with for the rest of your life.
I dreamed of writer but it was not feasible for me to write for living at that time. So I think I can write while studying following a goal: live as a researcher and die as a writer.
This can be a reason for writing this post right now, and a short story of the first step to become a researcher.
Getting an intern job
The first step to apply for an intern is to search for which company has fancy intern positions and meets your core values.
A simple process: apply with CV or cover letters, prepare for interviews and assignments, and pass all of them!
It can be a long story for me because the company I applied for has an introductory interview to get to know each other, an assignment, and a technical interview in the hiring process (longer for regular research scientists).
But stay with strong will and countless practice (and some failures). You can get where you want to reach finally.
Job interviews
What interviewers evaluate in the hiring process is closely related to what they care when they work.
For example, if one of their big concerns is now refactoring codes, they want you to show your ability to write clean and well-organized codes in the process.
For me, it was "clear interpretation of experimental results", "building models with evidence" and "clean code" not good model performance.
While doing your job assignments, it is possible to interact with interviewers or HR managers to ask your questions about the assignments, which I missed 😂.
Core values ✨
The smaller the company you want to apply for, the higher standards for core values you have to meet.
A startup I applied for focuses on AI research and encourages employees to be original and be nice to others. Wheter employee meets core values can be assessed during the hiring process and training period.
A core value: Self-motivation
Sometimes, growing company may not have concrete Roles & Responsibility.
It means that tasks assigned to you need to be designed by yourself rather than actually assigned by others.
While doing research, it becomes more important. But it is very hard for MS students or junior researchers to design and realize their own tasks.
Just ask for help!
Other professionals will help you with developing your idea. But don't expect them to draw idea from scratch because it should be your own idea. For me, it has been a common mistake to expect others to think rather than think for myself.
Organizational culture: Diverse and global backgrounds
There are many foreign employees in my company and the number of employees working abroad is increasing.
While working in a Korean company as a researcher, there are a plenty of times to read articles in English but no time to speak in English.
On the other hand, while working in a global company, there are many opportunities to interact with brilliant people from foreign countries or working abroad. It becomes a great chance to improve your research and communication skills. In my case, I attend meetings with foreign employees at least once a week.
In addition, it is also a great improvement with regards to "diversity". When you are working in a big company or involed in big organizaiton, it is likely to meet people who are like-minded or with similar (educational) backgrounds.
It is not always on the bad side, but it is always good to broaden your mind.
Dev culture: Code reviews
I haven't worked as an engineer for a long time but I've been always curious about what good system design is and what effect clean code has on deep learning research.
Don't take too long!
Learned from a short experience in a startup, there is no need for taking too long time to write perfect codes in a development stage, especially for code reviews (pull requests, PRs).
Short and functional codes with clear interpretation are enogh.
From a clean code perspective, it is better to split large (machine learning) system with small functional units, write unit tests if it's possible, and push them one by one. If it is inseparable, I don't think it is a well-organized system at the early stage and the system needs to be elaborated in design.
Closing
I wrote another post about applying for an intern abroad, which was helpful for Korean applicants preparing for a job in foreign companies. From that experience, I love to share my experiences and thoughts with others in similar circumstances.
Please leave comments if you have questions or concerns. Contact thorugh DM (LinkedIn) is also preferred.
Thank you for reading, bye! 👋