I've been doing the cosine similarity in plain ruby while I await the availablity of pg_vector on AWS RDS, scales well enough for small datasets for prototypes(

I've been doing the cosine similarity in plain ruby while I await the availablity of pg_vector on AWS RDS, scales well enough for small datasets for prototypes(talking less than 100 here, although haven't load tested it yet).