The amount of biological data being created today is mind-boggling. As a biologist or bioinformaticist, you probably know of places around the network that provide very useful resources for your task at hand -- but remembering the different ways to access this information is often a productivity drain. This article takes you through a step-by-step approach to building a Java-based Life Sciences Identifier (LSID) authority from scratch. It demonstrate how to build this on a minimal data set and on data downloaded from the protein sequence database Swiss-Prot, all on the Linux platform.