Return to Core
ReactAITailwindCloud Storage
AI Resume Analyzer
One résumé, one job description—structured feedback you can act on, not vague scores.
Problem
- Applicants get vague, un-actionable feedback; recruiters need consistent signal.
- PDFs are messy: you need extraction + structure before analysis is useful.
Build
- PDF processing pipeline (render + extract) feeding an AI step with strict JSON schema output.
- History/persistence layer for repeatable runs, review, and deletion controls.
Outcome
- Actionable, structured reports that render cleanly as UI components.
- A working live demo with saved analyses instead of a one-off run.
The thread
Built a resume-to-job matching workflow that converts PDF resumes into images, extracts text, and submits a structured prompt to produce ATS-style feedback.
Designed the app for reliable, schema-based outputs (JSON-only responses) so reports are easy to validate and render as UI components.
Added persistence by saving resumes, page images, and analysis reports to cloud storage, with history views and deletion controls for past runs.
Architecture Overview
- //React 18 + Vite SPA architecture with reusable UI components
- //PDF processing pipeline (pdf.js) for text extraction and page rendering
- //AI analysis step with strict JSON response schema validation
- //Persistent storage for analyses with history, view, and delete flows