Ask HN: Pivoting technologies in your projects?

In some projects there is some major difficulty, that is either solved or not solved. When not solved it often sits there until the end (maybe hidden behind some curtain). When it is solved, a solution might come from a new view on the architecture or some new technology, that plays the role of an enabler.

E.g. for a personal experience:

* Challenge: I tried to build a instant search web-interface over a large dataset with some non-trivial UI interactions.

* Things that did not worked well: jQuery, not because it's not nice, but because there were a couple of moving parts on the site.

* Pivoting technology: Using an Javascript MVC (knockout) made the interface slick and I could even implement features, that weren't on my schedule, because it was simple and manageable.

I am asking for experiences, because I observed a few teams finding such a pivoting technology in the past - and it's educational to see it in the context of their challenges.