AI Matchmaking & Recommendation Engine

About the project

Developing an intelligent matchmaking and recommendation engine for Hamqadam that transforms user profiles and explicitly stated partner preferences into meaningful, ranked match suggestions. Instead of presenting users with an unrestricted directory of profiles, the system identifies eligible candidates and ranks them according to mutual compatibility.Profile attributes and partner preferences are converted into comparable features through a custom feature-engineering pipeline. Eligibility rules first remove candidates outside a user’s defined requirements, after which a weighted compatibility model evaluates factors such as age, education, location, family background, and other selected preferences.The matching logic evaluates compatibility in both directions, helping ensure highly ranked recommendations align with the expectations of both profiles rather than only the person performing the search. Ranked recommendations are exposed through FastAPI services, with MongoDB managing profile and preference data.