Xooctory needs to address job distribution in a scalable and efficient way, by leveraging the available information to find the best agent to run a job.

This is a complex problem, and even though the best is not always possible, having an algorithm able to find a good one is already interesting.

To find the best agent, the system needs to know the resource requirements of a job, the available resources for an agent, and find a good match. For instance a frequent resource requirement is dependencies on artifacts (jars). Keeping a view of which artifacts are available in the local cache of agents, and which one are likely to be required by a build could help choose agents based on the size of required artifacts downloading.