After Finishing the Pre-Employment Part-Time Work
The other day, I finished my pre-employment part-time work.
I’ve experienced three departments, each for one to two months.
I don’t know my assignment yet, but I’ve summarized my growth points from this experience.
What I Learned
- Results first
- Problem-oriented
- Iterate quickly
Results First
Especially in ML/DS positions, you need to do everything from hypothesis testing to implementation.
Of course, there are cases where you’re only responsible for one part.
In these cases, you’re required to produce the desired results quickly.
A clear difference when transitioning from research to business is producing results.
Producing results—this is obvious in research too.
However, in research, the required results are often to exceed past accuracy using the latest methods.
In business, it’s different. If you can produce cost-effective results with classical methods, you should adopt them.
If you think the latest methods will produce better results, you need to verify and present to your supervisor.
Produce results. To do this, you need to draw lines between what to be particular about and what doesn’t need attention.
Problem-Oriented
The second is being problem-oriented.
What is analysis for? What use is the model you created?
If you forget this perspective, you’ll only focus on analysis.
What’s needed is raising business KPIs and clarifying and solving current problems.
Without thinking this way, you’ll waste time on self-serving analysis that doesn’t contribute to business at all.
My ideal engineer image is someone who can do everything from analysis to business proposals.
To reach that point, I need to be able to propose why I’m doing this analysis, what to clarify, and what measures or models to create.
Iterate Quickly
How quickly can you run the PDCA cycle? How much experience can you accumulate?
This is the shortcut to leveling up.
I believe how much I can challenge and fail while failures are tolerated as a new employee determines future growth.
With limited life remaining, I’ll focus on iterating quickly to produce results I’m satisfied with.
Finally
For students, I recommend trying internships or pre-employment part-time work if possible.
I was busy with research and am about to become a working adult without doing internships.
I have no regrets at all, but I think knowing early what’s required as a working adult is a good experience.
Behind the Scenes
I was reading both a presentation book and a data analysis book simultaneously, so I didn’t finish either today.
I think tomorrow and after will be reading records.
コメント