April 10th, 2007, 7:14 am
Alan's analysis is easily the most plausible.Certainly the advert implies either more than a little trouble in their thought processes, or some extremely state of the art arbitrage work.None of the technologies currently used by (or taught to) quants can do real time work on "complex text and numerical queries". Though if you throw enough resource at it, you can nearly get there.There exist technologies that will do this, and they do require a strong command of zero and first order predicate calculi. Sad thing is that most CS grads can't do this, at least not many under 30, and most quants knowledgeof Boolean logic doesn't get as far as the implies operator.If you don't know the tech I'm talking of then you can't do this job, unless it's some attempt to push vN architectures harder than is is reasonable. Would any CS grads like to guess what a vN architecture is ?It could be a good job, but won't be very quanty, indeed long long ago, when I first wandered into the markets I thought this was how "trading systems" would work, and I suppose part of me still does.There are reasons why they don't, even though we have a lot more power to punt at the problem.Sometimes I think the most valuable bit of s/w I ever wrote got thrown away. Was a bit of real programming that would acquire Bloomberg news data in close to real time, and shove it intolexical analysis code.Even though I mentioned it last on the forums nearly 3 years ago, I still get occasional PMs from people wanting it, and sounding quite sad when I tell them I left it at my last firm who did away with it.
Last edited by
DominicConnor on April 9th, 2007, 10:00 pm, edited 1 time in total.