Resumai - AI-Powered Resume Analyzer & ATS Checker
Upload a resume and job description, get ATS, tone, content, structure, and skills scores with concrete tips to improve.
Key Metrics
Problem
Most resume tools spit out a single vague score. Candidates still don't know whether they failed ATS parsing, buried skills, or mismatched the job description, so they rewrite blindly and apply again.
My Role
Solo product engineer end-to-end: UX, React Router UI, PDF ingest, AI analysis orchestration, auth/storage on Puter, and the results experience that turns scores into next actions.
Approach
Users upload a PDF (parsed client-side with PDF.js) and optionally paste a target job description. The pipeline scores five dimensions, ATS, tone, content, structure, skills, each 0-100 with 3-4 concrete tips. History is saved so people can iterate across versions. I designed the analysis view first (the money screen), then worked backward to upload and auth.
Tradeoffs
Using a hosted LLM shipped a useful product fast, at the cost of coupling quality and cost to the provider. I chose five readable scores over one mysterious overall number so users trust the feedback. Client-side PDF parsing keeps files off my server but limits exotic PDF edge cases.
FIG. 05 · FIVE SCORES
explainable beats opaque
In
Resume PDF
Optional job description
Out
Tips + version history
Analyze → revise → compare
Money screen · 0-100 each
ATS
Parse / keywords
Tone
Voice fit
Content
Substance
Structure
Layout clarity
Skills
Role match
Outcomes
- Public live demo with a full analyze → tips → history loop
- Role-aware feedback when a job description is provided
- Clear per-category scores recruiters and candidates both understand