SeamlessApply

The problem

Serious job searching means checking multiple boards daily, retyping the same answers into every application form, and writing a fresh cover letter per company — a process exhausting enough that most people apply to far fewer jobs than they should.

What I built

A platform built around the "Seam" — a saved job-search campaign with its own title, location, experience level, and target boards. Once a seam is running, the system scrapes matching jobs across four boards, uses GPT-4 to fill out application forms (through a companion Chrome extension that reads the live page), and sends a personalized cover email per job using the applicant's actual resume as context. A learning layer remembers how a user has answered recurring application questions before, so the same question never triggers a fresh AI call twice.

Everything runs two ways: reviewed manually inside the dashboard, or fully unattended through a daily cron job that applies the same matching logic the results page uses — one shared function, so the two paths can never drift apart.

Result

A working automation loop across four job boards with its own browser extension, form-fill pipeline, and cron-driven auto-apply — built solo from database design through to the extension's UI.