4 ms·

Ask HN: Project ideas for PhD student moving to industry

6 points by koots 8y ago

Hello, I'm finishing my PhD in CS in a couple of months and I'm applying for software engineering jobs with no luck. I get rejected from entry level jobs for being overqualified, and from more senior jobs for lacking industrial experience. I know C++, C, Python, a lot of algorithms and some ML, some data mining. I pass most of the hackerrank style interviews, but then I get rejections for not having enough experience. I have two months of writing and some other boring stuff. I can dedicate 10 hours a week for improving myself for job applications. Any ideas? Should I do a programming project? Any suggestions? I have interested in Algorithms, security, storage systems, ML, and DM. I appreciate your help.

sharemywin 8y ago
What's your PhD in CS subarea?(dissertation) Are there any jobs in that area? Why not teach?
- koots 8y ago
  My PhD is in Computer Science (mix of algorithms and data mining and some bioinformatics). I switched advisers few times. I can teach, I love teaching and I have huge experience being a teaching assistant. But there are some issues with teaching right now: 1-Most universities require teaching and research. 2-I want to move to software engineering work. 3-If want more and not start software engineering now, as a fresh PhD graduate, the problem will be worse later.
  - brutus1213 8y ago
    I think you are over glorifying software dev. Most devs don't do google/MSFT/Netflix-style SE. Most do the same crap we do in grad school except they have 100% more meetings and progress reports to fill out. When you say Bioinformatics, I assume your main skills are in Python? That is going to be a problem as a software dev unless you stick to data science/ML. Professional work requires expertise in Java (boring but necessary) and C++ (hard but efficient). Make sure you understand concurrency/parallelism/locking in your tools/languages.
    - koots 8y ago
      I know python and c++03. I agree that "concurrency/parallelism/locking" could be useful to me. May automotive companies ask for that and they also ask for some algorithms knowledge. Do you have any project idea in that area? Thank you for your help.
arandr0x 8y ago
Where are you located? You can apply to R&D roles (search for R&D engineer, research software engineer, R&D software specialist and similar titles) which are essentially software engineering roles that require grad school. A good idea may also be to pick a subfield that traditionally hires more educated people (science or engineering or chemistry software hires from that pool because the clients tend to be in academia). In a given city most of the PhD-friendly industry jobs are staffed by people who still regularly see your professors at happy hour. Have you talked to your professors and alumni of your program if they know anybody's who's in industry now? Look up on LinkedIn where the people from your lab ended up. You can probably get some of these people to give you an interview that's not biased against your specific experience.
- koots 8y ago
  In the middle east. They all continued to work in academia as postdocs. I don't want to go in that path. Especially that my most recent work was in a field I don't enjoy, but I had to do it because my previous adviser had to leave university and that was the only option or else I would have wasted the previous years effort in vain. I have lots of theoritical experience in many areas and I just need to start in a first job. And from there it would be easy. As I will be able to say I have actual industrial experience. I want a software engineering job. I even apply for entry level jobs and I get rejected. I even got rejected several times for internships.
  - arandr0x 8y ago
    Do you get to the interview, or they don't even reply to your resume? Because if you get interviews you need interview practice not projects. Assuming you do have, like, school projects, and your PhD project includes programming. Otherwise you do need projects. For C++ a good way to have projects is to talk to PhD students in other fields (physics and bio are good) and ask if they need help with their computation and modeling. Listen to them a lot as they often already use software for most important things but may have need of e.g. quickly indexing text files or targeted data visualizations. In industry, we like hearing about collaborative projects.
koots 8y ago
Should I interpret not getting response about my question for a project idea as a sign that it is not a good idea?
- brutus1213 8y ago
  I think the other commenters suggesting a better role is the best idea. Believe it or not ... there is a lot of industrial R&D ... people don't publish papers on all of it. Key outputs are Proof of concepts and patents. If you find a role like that, you will likely have the best chance to succeed. That said, lets talk about the project idea .. here is what I suggest. Do a solid project (or three) and give a talk on it at a meet up or tech conference. Data science/ML is stupid hot right now as an area. But the business people don't give a crap about algorithms. They want outcomes or whatnot. So, when you do a project, don't go fully deep into an algorithm or eval. Focus on a business problem (could be some trivial data analysis on open data) and write a blog post on it/do a talk. Use visualizations. use tools used in industry for this (e.g. Tableau or whatever). Best of luck to you.
  - koots 8y ago
    I agree. But all data mining/machine learning projects I can think of are either trivial or typical. That is why I was asking for suggestions. Thank you.
    - arandr0x 8y ago
      The Kaggle competitions are good projects, just don't expect to win. I like it because I don't like studying on toy stuff. I hear it can be a social thing for some too so maybe you will meet people in your area that know of jobs for PhDs.
      - koots 8y ago
        Good idea, I will check it out.

6 points by koots 8y ago

sharemywin 8y ago

What's your PhD in CS subarea?(dissertation) Are there any jobs in that area? Why not teach?

koots 8y ago
My PhD is in Computer Science (mix of algorithms and data mining and some bioinformatics). I switched advisers few times. I can teach, I love teaching and I have huge experience being a teaching assistant. But there are some issues with teaching right now: 1-Most universities require teaching and research. 2-I want to move to software engineering work. 3-If want more and not start software engineering now, as a fresh PhD graduate, the problem will be worse later.
- brutus1213 8y ago
  I think you are over glorifying software dev. Most devs don't do google/MSFT/Netflix-style SE. Most do the same crap we do in grad school except they have 100% more meetings and progress reports to fill out. When you say Bioinformatics, I assume your main skills are in Python? That is going to be a problem as a software dev unless you stick to data science/ML. Professional work requires expertise in Java (boring but necessary) and C++ (hard but efficient). Make sure you understand concurrency/parallelism/locking in your tools/languages.
  - koots 8y ago
    I know python and c++03. I agree that "concurrency/parallelism/locking" could be useful to me. May automotive companies ask for that and they also ask for some algorithms knowledge. Do you have any project idea in that area? Thank you for your help.

arandr0x 8y ago

Where are you located? You can apply to R&D roles (search for R&D engineer, research software engineer, R&D software specialist and similar titles) which are essentially software engineering roles that require grad school. A good idea may also be to pick a subfield that traditionally hires more educated people (science or engineering or chemistry software hires from that pool because the clients tend to be in academia). In a given city most of the PhD-friendly industry jobs are staffed by people who still regularly see your professors at happy hour. Have you talked to your professors and alumni of your program if they know anybody's who's in industry now? Look up on LinkedIn where the people from your lab ended up. You can probably get some of these people to give you an interview that's not biased against your specific experience.

koots 8y ago
In the middle east. They all continued to work in academia as postdocs. I don't want to go in that path. Especially that my most recent work was in a field I don't enjoy, but I had to do it because my previous adviser had to leave university and that was the only option or else I would have wasted the previous years effort in vain. I have lots of theoritical experience in many areas and I just need to start in a first job. And from there it would be easy. As I will be able to say I have actual industrial experience. I want a software engineering job. I even apply for entry level jobs and I get rejected. I even got rejected several times for internships.
- arandr0x 8y ago
  Do you get to the interview, or they don't even reply to your resume? Because if you get interviews you need interview practice not projects. Assuming you do have, like, school projects, and your PhD project includes programming. Otherwise you do need projects. For C++ a good way to have projects is to talk to PhD students in other fields (physics and bio are good) and ask if they need help with their computation and modeling. Listen to them a lot as they often already use software for most important things but may have need of e.g. quickly indexing text files or targeted data visualizations. In industry, we like hearing about collaborative projects.

koots 8y ago

Should I interpret not getting response about my question for a project idea as a sign that it is not a good idea?

brutus1213 8y ago
I think the other commenters suggesting a better role is the best idea. Believe it or not ... there is a lot of industrial R&D ... people don't publish papers on all of it. Key outputs are Proof of concepts and patents. If you find a role like that, you will likely have the best chance to succeed. That said, lets talk about the project idea .. here is what I suggest. Do a solid project (or three) and give a talk on it at a meet up or tech conference. Data science/ML is stupid hot right now as an area. But the business people don't give a crap about algorithms. They want outcomes or whatnot. So, when you do a project, don't go fully deep into an algorithm or eval. Focus on a business problem (could be some trivial data analysis on open data) and write a blog post on it/do a talk. Use visualizations. use tools used in industry for this (e.g. Tableau or whatever). Best of luck to you.
- koots 8y ago
  I agree. But all data mining/machine learning projects I can think of are either trivial or typical. That is why I was asking for suggestions. Thank you.
  - arandr0x 8y ago
    The Kaggle competitions are good projects, just don't expect to win. I like it because I don't like studying on toy stuff. I hear it can be a social thing for some too so maybe you will meet people in your area that know of jobs for PhDs.
    - koots 8y ago
      Good idea, I will check it out.