Although it seems to us in 2021 as though scientific high-performance computing (HPC) has been around for a while, it's actually a fairly young field. As a matter of fact, it's several young fields—computational astrophysics, computational chemistry, computational genomics, high-performance machine-learning, uncertainty quantification, and many more. These fields, on close examination, have not really stabilized or optimized their collaborative processes in a manner analogous to that of more mature, "classical" sciences. As a consequence, valuable science is often needlessly lost, or left uncollected. I believe that this situation is likely to change as these fields grow up, and assimilate or adapt management practices characteristic of mature collaborative sciences. In particular, a new key role is likely to be formalized in HPC collaborations, analogous to the lab manager of chemistry labs. In this essay I will refer to this role as the operations manager.