Perhaps the most valuable achievement of a highly-functioning Knowledge Management system is the ability to identify a colleague within your firm who has pretty much the exact expertise you're looking for, when you need it.  I call this the "Ask Sally" moment, as in "Ask Sally; she'll know." 

Traditionally, firms that have tried to create this capability have approached matters with a fairly blunt instrument:  Surveying everybody, or at least every professional, to ask them point-blank where they consider themselves an expert.  Of course there are any number of problems with this, from the practical (it takes time; it needs constant updating) to the epistemological (do you mean what I mean by "expertise?").  Stories of companies investing millions to create such systems, and then watching them lie dormant and neglected, are legion.

But what if it turns out that your firm might already know most of what it needs to about your professionals?  If you cobbled together—this means you, IT!—information from many of the internal databases you already have, might you not end up with a reasonable facsimile of such a system? 

For example, HR has information on everything from what office and department you're in to where you went to law school, what CLE topics you've studied, and who you've worked with (performance reviews).   Finance and accounting know which clients and industry groups you've worked for and how much and for how long (time and billing records).   Marketing knows if you have any articles, whitepapers, or even patents to your name.  &c.   Pull all these together with baling wire code, set an internal version of Google or Verity loose on it, and you might be surprised how far you get.  This could be a case of knowing more than you think you do. 

Better yet, you don't have to survey anybody, and the information is continually refreshed through the ordinary course of doing business.  Before you give me too much credit for this, read McKinsey's take on it.