Milking the Cash Cow: how to make tons of money by keeping millions of lines of Erlang code running.
Klarna was founded in 2005 (as Kreditor), with little more to its name than an idea and a rather hastily assembled Erlang program. In 2014 that program, kred, handled several billion euro, generating a revenue of 160 million euro. Like most code that has been through multiple generations of developers, it is generally disliked by its current caretakers. At the same time, it’s amazingly featureful and reliable. This talk will be about kred, and lessons learned about how to wrestle such a beast through it’s various stages of life.
Entertain and educate about working with business-critical legacy code.
People interested in stories about code.
Mats holds what’s possibly the least useful of all academic titles: a Ph.D. in Experimental Nuclear Physics. He’s spent a good amount of time chasing exotic subatomic particles, but rarely found any interesting ones. Since 1997 he’s been paid to chase Erlang bugs instead, and has been able to find quite a few, possibly because they are not hard to come by. He currently works for Klarna, architecting Payments, Next Generation.