I think this has become a cliché here (a true one, however).

The interesting question is what sort of jobs contain a lot of challenges that are technical in nature.

The interview process is where the challenges are technical in nature. The only time I have been asked to implement a sort algorithm after university was for interviews.

Generally I find that many companies have a small amount of difficult problems. The trick is to become the person who gets to solve them when they come up.

At one of the places I worked, I got to implement a request batching algorithm using dynamic programming, at another place I got to implement a nurse rostering algorithm (this is not as easy as it sounds, and is an active research area), and at a third place I implemented the recursive tree algorithm that keeps a lookup table of user permissions up-to-date in a situation where roles inherit permissions from other roles.

Holy shit, this.

Let me recommend to you "Peopleware: Productive Projects and Teams" by Tom DeMarco & Timothy Lister (go with the 3rd edition) and "The Mythical Man-Month" by Frederick P. Brooks Jr. (go with the anniversary edition). Both books expound the idea that most software projects fail not because lack of technology or innovation, but rather due to the mismanagement of people and a lack of understanding how teams work.

Anything specific about the 3rd ed of “Peopleware” as opposed to the 2nd? Thanks

I haven't read the 2nd one, so I should have just said that read the latest that you can. The main contents can feel quite dated for any edition, but there have been updates to fit the times that I'm not sure are in earlier editions than the one I read (i.e. the 3rd edition).

Added to https://github.com/globalcitizen/taoup

Agreed, learned this over the past couple years while in management. Its usually a process issue or something related to keeping an engineer happy.