The only fields that are shared between QC and customer data are assay name, J numbers, and lot numbers (Table 1). As each analyzer has a unique J number, we used this shared information to match QC samples with customer data. Although this approach works in most cases, there are instances when the same customer processes multiple samples, potentially from multiple analyzers, but logs only one call. Thus, the data that will be used to train the predictive algorithms are, in essence, noisy.