Everything I have written that is tagged Infrastructure.
A visual case for local AI: cached context, busy agents and batching could turn faster workstations into much cheaper inference.
