think-cell

Description

Hiring process reviews

• “They tried to evaluate at extreme level. I think the details they need can be evaluated in better way by using different methods optimally “
  
• “Yuliya made it very easy.“
  
• “Not great, at all, on technical interview “
  
• “Everything went professionally and so clear.“
  
• “I couldn't relate the online test queries with the way a Software tester normally work. Things were half in queries and we are required to find a solution for that. Google was an option left and I tried to understand the issue using the same. Then I provided an answer in the way they said clearly, even then I wasn't selected mentioning that it didn't meet their standards. Only if standards and measurement criteria were clearly, we will be able to perform well. I am very disappointed about the test.“
  
• “The assignment was not at all related on the industry test practices and application rejection feedback was not shared.“
  
• “The test question at think cell was not clear “
  
• “They have a very fair, but tough, coding challenge. I liked it, even though I couldn't find the right solution.“
  
• “I like the way the HR from the Thoughbyte approached me about an open position. They also continuously were in contact with me throughout the process. Appreciate the quick replies.“
  
• “It felt a little bit disrespectful, to invest nine hours of your time doing a task for a position to be auto rejected instantly by an automated process for a small detail. In the place I live I was only once tasked by a full day trial, and as a sign of appreciation after throughly examining your work they would arrive to a conclusion. Also if accepted of not for the trial they invited you to lunch at their facilities during the interview and gave presents for candidates for their invested time either if accepted or not just as a sign of apreciación. The feeling that you get after doing a task for think cell is totally the opposite and feels that they even bother to seriously bother to take a look at your solution, let alone try to make an effort and show a gesture of good will that they appreciate your time spent on trying to get accepted after being approached by them for a job.“
  
• “I think the idea is good, however the implementation with the test is lacking. The goal would be to eliminate the lower skilled programmers from the start so they don't waste their time with in-person interviews, which is great. But for this to work, the test needs to be actually good and have a wide range of metrics so that you can separate candidates in a granular way. Instead the test has a binary result which may not reflect the actual quality of a candidate. I also think there are issues with the test itself, it's nice that you get two chances but the error you get if you submit a bad solution should be more expressive. We work with compilers, if you mess something up, you get an error about it and you can fix it. In my case I failed twice the data type constraints, which in first case was absolutely correct, but in the second attempt I can't figure out what "illegal" operations I was triggering, since I tested with data types created based on the requirements. Either the tests are off, or the problem description was missing key information. The problem itself is not difficult, the difficulty is given by "gotcha" style requirements that are designed to drive the interviewee towards incorrect solutions. As a side note, I searched parts of the problem on the web after the test and there are quite a few threads on sites like github/stackoverflow about this, so the message about "can't give feedback" is invalid. In the end, I think the idea of it is nice but could be improved. If the problem is a glimpse in the kind of work is to be expected at think-cell, I'm happy I failed. I would hate to have all the requirements incomplete and no feedback from the product owners on what they actually need.“
  
• “The task had average difficulty (at least i thought so :) ). No feedback on what went wrong with the code. I tested my code before i submitted it and the response was it failed on the correctness test. At least, they could have given me the test data on which correctness test failed...“
  
• “Great support and guidance from Alina .“
  
• “regarding the online test, really difficult to access what company is expecting from our answer. Though the test was difficult, the interview process was smooth and professional.“
  
• “It was really interesting coding challenge.“
  
• “This company's hiring practices are questionable, to say the least. Despite claiming to seek top-tier talent, they insist on using outdated textbook interview questions. What's more, over half of the test cannot be edited within their browser, and they rely on a five-year-old compiler version. Ironically, they complain about ignoring numerous applications after candidates submit the test, all the while overlooking the fact that the uneditable portion is the main issue. To my fellow C++ developers, there are better opportunities elsewhere where your well-honed skills will be truly appreciated, rather than being reduced to a single test.“
  
• “Only 2 attempts of submission to a very complicated 9-hour long problem. Submitting a pull request to open source repositories or coding competition can have multiple attempts.“
  
• “the text of the requirements is not correct and very confusing. “
  
• “The coding test is not practical enough for the position“
  
• “The algorithmic task is well-thought, as it requires the candidate to reason properly about the given data structure, and also requires a good knowledge of how the standard C++ library works. On the other hand, the running time requirements are simply unfair. First of all, what does it mean "Big-O optimal"? No computational complexity or algorithms book ever mentions this notion. When dealing with complexity, first one has to clarify if you are referring to worst-case or best-case analysis. And even if this is clarified, your notion of optimal is unclear. If you mean asymptotically optimal, and thus you mean that the running time in the worst case, or best case, depending on what you want, is both O(f(n)) and Omega(f(n)) for some function f of the input length n, then there is no way for your automatic system to asses this is the case, and moreover, my solution was asymptotically optimal in that sense. However it is clear you are looking for a solution that essentially matches the one you have, which I fnd unfair. The fact that your solution employs a few less O(log n) function calls does not make it necessarily faster than other implementations, as this highly depends on the structure of the given input and other underlying assumptions. Be more formal and clear on the running time requirement, or simply ditch it altogether, as it is simply broken in the way it is stated.“
  
• “The interviwe was ok But the test that was givien was not, im not talking about the difficuly of the problem but that it was nit understood. If there was more examples about the input and output then it will be fine i think “
  
• “Misleading specifications and problematic timespan assertion.“
  

Questions