Rice University logo
Top blue bar image
Computer Scientist at the Department of BioSciences, Rice University

interview notes – Houston tech company

I interviewed (via phone) for a Software Developer position at a Houston tech company

The company:

  • Very into cloud and other new technologies
  • Has grown from 300 to 1,000 employees in the last two years

The questions that I want to remember:

  • How to handle memory problems in Java when a program is supposed to run for a long time (weeks or months)
    • I did mention (among other things) that when an object is no longer going to be used, we clear it via setting null whatever references to it.
    • I should have also mention that the ‘finalize’ block could be used when defining classes to clear up resources
  • How do you use design patterns
    • I only mentioned a couple, but I actually use several more that I did not remember at the moment.
  • Unit testing
    • I have done some but not much. Here I should have added that some functionality can’t be tested automatically, such as image acquisition from microscopes, User Interfaces.
    • I should have mentioned that some of the statistical tests I implement, are checked with the result that comes up via Excel. When they match, then my code will work OK with not just two sets of data but with all data in a particular database (such as tens of thousands of records).
  • Last, I actually dislike the questions along the lines of “How do you see yourself 5 years from now?”
    • Looking back, I know things I would not have imagine that I would know 5 years ago (because I have been in a genetics lab) but they are not directly relevant to an IT company
    • On the other hand, there is a set of technologies that have not truly changed. It seems that each interesting technology from 5 years ago has gotten more projects around it and become more diverse.
    • The exception is probably research: not as much growth in the diversity of Computer Science research in a decade.
      • There may be a few notable new things but not many.
      • The usual is “new type of data?”, then new algorithms for it. Just see how many research articles use twitter data

Leave a Reply