The code challenge was manageably hard, but it required a lot of experience with data structures, like they will let you know in advance. The major issue is that there is no context given for the code you're implementing, so it's a little hard to understand what exactly the requirements are. Since there is no communication with the company during the exam you can't ask questions that you normally could in your workplace to get a better understanding of what you're doing and trying to achieve. When you submit your code for automated review and it does not meet the requirements you get one more chance to fix these. The automated response will tell you what requirement your code failed to fulfill. The problem with this is, that it will not show you all the requirements your code doesn't fulfill - if you fail the first requirement, then it will only mark that requirement as failed, but leave you with no information about whether or not your code fulfills the others until you submit your fixes for the first issue. This is certainly something that should be improved upon by think-cell, but also wouldn't be an issue if your code fulfills all the requirements from the start. It's very generous of them to give you a second chance.