ApplyMate

Project Overview

A native Android app built in Kotlin with MVVM. Photograph or paste a job posting and ApplyMate reads it with on-device Google ML Kit, extracts the deadline, contact details and required skills, scores the posting against your own skills, and saves it as an application you can track from Saved to Offer. Built as a group project for Mobile Development Advanced at Fanshawe College.

Demo Video

Key Features

• Scan a posting from a photo or screenshot (Text Recognition v2) or paste its text
• Entity Extraction finds the email, phone number, website, address, and deadline automatically
• Skill match score comparing the posting against your skills, listing the ones you are missing
• Application tracking from Saved to Applied, Interview, Offer, or Rejected, with notes and a stats dashboard
• Detects non-English postings and translates them on the device; Smart Reply suggests responses to recruiters

Technical Challenges & Solutions

Matching skills accurately in free-form posting text, so "Java" does not match "JavaScript" while "REST APIs" still matches "REST API". Solved with a plain-Kotlin SkillMatcher using whole-word matching against about 45 common skills plus the user's own, covered by unit tests. All ML runs on the device, so after a one-time model download the app works offline.