First  we have each person in an existing relationship answer survey questions  plus assess the quality of that relationship. Above couple P25/P26 has a  great relationship (96% satisfaction) while couple P47/P48 has a poor  relationship (29%). Then we train a machine learning engine to recognize  patterns in the survey answers that predict quality of relationships.  For an example question “I am religious”, someone answering “Strongly  Agree” may not be a good match for someone answering “Strongly  Disagree”. For the question “I want to have children”, both answering  “Strongly Agree” would be a good sign. 

Finally,  for someone (A) seeking a relationship we ask those same survey  questions and the trained machine learning engine assesses them with  answers from others seeking relationships. Pairing A/B76 doesn’t look so  good (56%) but pairing A/B65 looks promising (95%), so we recommend  candidate B65 to person A.