[Case Study]

PairUp

Developed PairUp, a full stack MERN application that utilizes React.js and Typescript on the front end to create clean, versatile, and reusable components. Integrated a Node.js backend that connects to an embedded MongoDB database to perform semantic search using vector indexes.

Overview

Problem

Keyword search treats a resume as a bag of strings. Recruiters end up sorting through false positives and missing strong candidates whose vocabulary doesn't match the JD.

Approach

React + TypeScript front-end with reusable Shadcn primitives.
GraphQL via Apollo for typed end-to-end contracts.
MongoDB Atlas vector index over candidate embeddings; structured filters layered on top for explainable ranking.
Stripe Checkout + webhooks for the paid tier.

Outcome

Live beta. Semantic match scoring across thousands of profiles, paid conversion working, and a UX recruiters actually trust.