PIPELINE ARCHITECTURE

How the AI Pipeline Works

From raw job description text to an ATS-optimized, fact-checked resume PDF in four automated steps.

01
RAW TEXT PARSING

Input Job Description & Resume

Paste the target job description alongside your current resume into our dual-pane editor. The Gemini pipeline extracts structured requirements, must-have skills, and resume work history.

  • ✓Identifies core technical skills & soft requirements
  • ✓Extracts company names, dates, titles, and bullet lists
  • ✓Maintains total data privacy inside your browser session
02
TRANSPARENT MATCHING

Deterministic Skill Diffing

The system compares your candidate skills against the JD criteria using normalized token matching (handling plurals like APIs ↔ API and acronyms like SSR ↔ Server-Side Rendering).

  • ✓Calculates exact Match Score percentage (0–100%)
  • ✓Generates git-style + Matched and - Missing skills diff
  • ✓Highlights matched skill pills in your reordered skills list
03
FACT-CHECKED REWRITING

Anti-Fabrication Tailoring

The LLM reorders and rephrases your existing experience bullets to prioritize JD-relevant achievements. Built-in validator guardrails automatically reject any attempt to invent new dates, titles, or metrics.

  • ✓Reorders bullet points to lead with role-relevant keywords
  • ✓Single-retry validation loop blocks invented facts
  • ✓Preserves exact company names, titles, and date ranges
04
ONE-CLICK DOWNLOAD

ATS-Safe PDF Export

Generate a clean, single-column PDF resume using @react-pdf/renderer. Formatted specifically for optimal parsing by ATS recruiting software like Lever, Greenhouse, and Workday.

  • ✓Single-column ATS compliant layout
  • ✓Clean typography and standard section hierarchy
  • ✓Instant vector PDF download ready for submission

Try it with your resume now

Test the pipeline with any job posting and see your instant match score.

Launch Tailoring Tool →