DataRepublican (small r)@DataRepublican
I'm working on ActBlue donor search and need to rethink the index structure due to the sheer number of unique names (i.e., unique tokens). Unlike standard English words, which number around 50,000, names have far more variations, making efficient indexing more challenging.
This wasn't an issue for principal officer searches because my database contains around 100,000 nonprofits—manageable in comparison. However, ActBlue donations involve over 25 million records, which significantly increases the complexity.
That said, this isn’t an insurmountable problem. I should have a working, searchable solution by tomorrow.