[rch]

I'd like to come at it from the opposite direction, and help engineers build out profiles of what they are looking for. Not another resume site, but a forum for making anonymous declarations of what the the perfect job(s) would look like. And something similar for potential employers. I probably wouldn't even focus on making connections, but instead just provide scores and feedback on availability, trends, regional factors, indirect costs, etc.

It seems like that sort of data would be worth something to new and established businesses, who both have to sort through an increasingly crowded forest of tools and technology.

[humanrebar]

Isn't this sort of what LinkedIn is for?

[logfromblammo]

It's what the theory of LinkedIn is for. In practice, LinkedIn is a cesspool of useless recruiter spam, from the very people that drove the employers and candidates to try LinkedIn in the first place.

It was destroyed well before it ever became useful to me. I won't even try the next iteration--whatever it's called--until I am convinced that they can keep the bozos out.

[rch]

I find linkedin to be relatively useless in this context. I need to be able to supply scores and weights to the various things I'm looking for, and to submitted job descriptions. I'd like to get at what candidates are looking for, not just what they have done.

It makes sense to list Java and C# in a profile, but I'd give a bump to Python and Haskell opportunities, personally (others might do the opposite).

Actual hiring is driven more by accomplishments anyway.

[vonmoltke]

What you are describing sounds more like OKCupid for engineering jobs.

[rch]

Maybe e-harmony, since it'd be double blind model-driven matching, but with near real time feedback while one is refining a set of preference parameters.