I have a brief introduction to this subject in another page here, where I make the point that the help desk paradigm provides the classic practical example. However, my rather broader view of the knowledge management field allows me to see a great many more examples that, if I have anything say in the matter, should also be regarded as equally classic.
To put a fine point on it: I consider modern intranet-based systems that facilitate such functions as travel requests, expense submission and approval, time-reporting and approval, general purchasing request and approval to be operate in the knowledge management domain. These apply to organizations. Then there are those that facilitate the business processes of software engineering firms, such as code repositories and build management, in-house issue and feature tracking, virtual farm management, etc. All of these systems - which are borne by some kind of technologically-based infrastructure - both facilitate work processes and leave a trail that not only allows anyone to see what others have done, but to collaborate in ways that would otherwise be impossible.
I've lately developed another angle to knowledge management - namely, product knowledge management (PKM). Whereas knowledge management is about improving the performance of an organization for the organization's own sake, PKM is about doing it on behalf of customers. The idea is to integrate the creation and conservation of product information throughout the product development process to optimize the use of resources needed to explain complex products to their users.
There's another species of knowledge management that is less technological than cultural, depending more on learned behaviour than work-process necessity. The best example I have of this at this moment is the community of practice (CoP), which is the modern analog of the traditional guild or apprenticeship model. The fundamental purpose is introduce/expose entrants to a knowledge domain to the work of experienced practitioners as a matter of how the work is conducted. This doesn't obviate the need for, or eschew formal training programs, but changes the way work is conducted to legitimize the participation of less experienced practitioners.
Etienne Wenger has done considerable academic and practical work in communities of practice. Also see the site he shares with Beverly Traynor and their discussion of communities of practice specifically.