Offerly

A job hunt is a project management problem wearing a career costume — dozens of applications, endless resume variations, and no single place to keep it all straight. The product needed to make the whole search feel organized and give the seeker an unfair advantage on every application.

1. 01

   Built a kanban tracker so applications move visibly through stages, giving the search a sense of momentum and control.

2. 02

   Added AI resume tailoring against a specific job description, plus cover-letter and interview-prep generation, driven from a reusable “career fact library.”

3. 03

   Scored resumes for ATS compatibility so users could optimize before submitting, not after being filtered out.

4. 04

   Let users import GitHub repositories as portfolio projects and export or back up all of their data at any time.

Offerly is the productized rewrite of an earlier local, single-user app that relied on a bridge process running on the user’s own machine. The migration — captured in a heavily consensus-reviewed plan — lifted every feature into a hosted, org-scoped SaaS with a three-tier authorization model and per-org bring-your-own-AI-key support. The design variety lives inside the product rather than in competing app shells: six web resume previews and eight distinct LaTeX PDF templates (Bauhaus, Chicago, Harvard, Oxford and more), composed with four brand palettes across light and dark.

This began as a scrappy single-user tool, so the rebuild was a genuine collaboration rather than a fixed brief. Every major decision — hosted multi-tenancy, the three-tier permission model, bring-your-own-key AI — was talked through directly and captured in a plan we reviewed together before I wrote any code. You could reach me with an idea and see it reflected in the product days later, not quarters. That direct line is what let the scope grow confidently without ever losing the thread.