you took 5 years to decide econ research was not for you?
A lot of people decide research is not for them even after getting an AP job. It is perfectly fine for that, your research experience will change throughout your PhD study.
Finishing econ PhD wants to break into software engineering/ML
-
you took 5 years to decide econ research was not for you?
At first, I thought it was an interesting subject. But as you dive deeper into the forefront of empirical research I think much has been done and academics resort to using secret data for story-telling, gender/diversity studies in labor turns me off real hard. I also wondered why modeling macroeconomy can't be more realistic, i.e. Ditching representative agent models like DSGE and simulating an economy with an agent-based model, this would be a good candidate in leveraging recent advancements in computing power
Anyway, I don't mind starting from fresh even from an intern, junior role as I have lots to learn. The closest job would be as a quant but as others said if you don't know anyone in the field it is hard to get job, adding to that they hire mostly Physics, Math grad.
I am interested in game development too so my to-do list after C++ is OpenGL and perhaps UE5.
Is the market very saturated in software engineering? For machine learning, it seemed the hype has died down, and to be honest other than FAANG most companies still rely on ancient models like SVM.
-
You can't be a "software engineer", that is much too broad, you need to focus on specific tasks and/or languages/systems, for instance scientific computing on Unix/C, or video games using Unreal/C++. My advice echos some of the above, first find an economist/data job, and in the meantime work on a open source project close to what you want to do. This will give you the experience and credibility you need for a developer position.
-
I have completely lost interest in econ with the current state in academic research…
Finishing PhD in a few months and would I be too late to break into junior software development/engineering/data science? Any transferable skills?
The reason is I am hooked with game security testing and manipulating game address needs lower level language like C++ and I am learning that…
During research and as hobby I know python enough to write web crawler in Selenium, ML pytorch like pytorch. R with data wrangling like dplyr, data.table etc…
Ideas?not sure if you will be able to get an interview at tech companies. they used to hire PhDs from other fields like chemistry and physics. But is getting much much more difficult in recent years for you to get an interview as a new grad if you did not take enough CS related classes. Once you get the interviews, things are much easier as you only need to ace leetcode problems.
You're off to a good start
My advice is to take about two months and really focus on LeetCode. Seems like Java is the preferred language these days as everything is web-centric though coding solutions in C++ is always impressive as well.
Data Science is a crowded space with a lot less end-market demand than SWEs. If you want to do DS, I would suggest going through your university's placement center in terms of contacts to set up informational interviews. Without contacts and previous experience, you're stuck trying to do an independent entry level job search.Beware, OP, the above is mostly trõlling. Nobody uses Java these days. JavaScript, maybe, if you're in web frontend development, but that's not likely to be a position you're aiming for anyways.
Python and R are the only relevant coding skills (besides SQL, but that isn't actual coding).
-
I would not also underestimate the amount of effort, study and frustration that effectively retraining from scratch involves. You will need, at a minimum, proficiency in one programming language. Something like C++ Primer is 1000 pages alone, and while all of it is important, it is just "tools" just like a year-long econometrics sequence is just tools. Learning how to use all of that in real projects will require yet more reading and countless hours of trial and error. You also need to become familiar with frameworks/libraries for your target language/system. This may be something like Qt for GUI development, or a video game engine, or some JS Web framework...
-
Don't. Swe jobs are completely commoditized and most of the time they are mind numbingly boring.
Find some rent seeking job leveraging your PhD. In Swe, learning never stops, you always have to study the next hot tech or the next the framework. When you are young, that's fine but it takes its toll on you as you get older. You really don't want to compete a/u/dists that work 15 days a day, either.
-Actual SWE
-
Don't. Swe jobs are completely commoditized and most of the time they are mind numbingly boring.
Find some rent seeking job leveraging your PhD. In Swe, learning never stops, you always have to study the next hot tech or the next the framework. When you are young, that's fine but it takes its toll on you as you get older. You really don't want to compete a/u/dists that work 15 days a day, either.
-Actual SWE^ Software developer is hugely overrated, long hours, not as much money as you think, boring, commoditized work.
-
Don't. Swe jobs are completely commoditized and most of the time they are mind numbingly boring.
Find some rent seeking job leveraging your PhD. In Swe, learning never stops, you always have to study the next hot tech or the next the framework. When you are young, that's fine but it takes its toll on you as you get older. You really don't want to compete a/u/dists that work 15 days a day, either.
-Actual SWE^ Software developer is hugely overrated, long hours, not as much money as you think, boring, commoditized work.
This. Earnings and job prosects were sweet until 10 years ago, but talent supply is catching up. The field is good only for the senior SWEs now. For a newcomer like OP, it is better to levarage that Finance PhD
-
Thank you all,
Indeed knowing Python or R != knowing C++ it is several times more complicated and downright evil.
For example concepts like compile-time vs run-time, linker, const/constexpr, types-declaration/conversion, references/pointers, bit management don't exist in those interpreter languages so don't say you know to code until you try C++!
Java has a similar syntax to C++ from what I heard, the only difference is Java does types-declaration/conversion automatically.
Joining a local meet-up is one way to get to know them this is good advice!
-
programming is easy
I would not also underestimate the amount of effort, study and frustration that effectively retraining from scratch involves. You will need, at a minimum, proficiency in one programming language. Something like C++ Primer is 1000 pages alone, and while all of it is important, it is just "tools" just like a year-long econometrics sequence is just tools. Learning how to use all of that in real projects will require yet more reading and countless hours of trial and error. You also need to become familiar with frameworks/libraries for your target language/system. This may be something like Qt for GUI development, or a video game engine, or some JS Web framework...
-
15 days a day
Don't. Swe jobs are completely commoditized and most of the time they are mind numbingly boring.
Find some rent seeking job leveraging your PhD. In Swe, learning never stops, you always have to study the next hot tech or the next the framework. When you are young, that's fine but it takes its toll on you as you get older. You really don't want to compete a/u/dists that work 15 days a day, either.
-Actual SWE