User Provides Job Posting
The process begins with a URL to any job posting or pasted job description text. Both paths are fully supported.
- Job board URL (Indeed, LinkedIn, Oracle HCM, Workday, etc.)
- Pasted job description text
Two supported input paths
Job Description Extracted & Saved
The full job posting is parsed and saved to the knowledge base for reference throughout the process.
- Job title & company name
- Location & work schedule
- Required and preferred qualifications
- Key responsibilities and action verbs
- Industry-specific terminology
- Application deadline
Career Data Loaded
All relevant career documentation is loaded to inform the tailoring process. This data is maintained across sessions.
- Master resume (base document)
- Full career history with project details
- Dated achievements log
- Performance appraisals and recognitions
Targeted Questions Asked
Before tailoring begins, a focused Q&A captures any new or role-specific context not already in the knowledge base.
- Any new achievements or certifications since last update?
- Specific experience particularly relevant to this role?
- Skills or tools to emphasize for this position?
Job Keywords Extracted & Weighted
Keywords and phrases are extracted from the job posting and weighted by prominence to prioritize alignment.
- Title / header keywords → 3× weight
- Required qualifications → 2× weight
- Preferred / nice-to-have → 1× weight
- Responsibilities & action verbs → scored
Resume Tailored to the Role
The resume is rewritten to maximize keyword alignment and role relevance. Every change is grounded in real experience.
- Professional summary rewritten to mirror job language
- Skills section reordered by relevance to this role
- Experience bullets reframed to emphasize matching responsibilities
- Quantified achievements preserved and prioritized
Nothing is fabricated. Every statement is sourced from documented career history. The AI tailors emphasis and language — it does not invent experience.
Match Score Calculated
A keyword overlap score is computed to measure resume alignment with the job posting.
- Formula: Matched keywords ÷ Total job keywords × 100
- Output: Match % score
- Output: Full list of matched keywords
- Output: List of gap keywords (missing from resume)
PDF Generated via Professional Template
The tailored resume is rendered into a polished, print-ready PDF using a custom HTML/CSS template optimized for hiring manager scan patterns.
- Two-column layout — certifications and skills on the left margin (first thing eyes scan)
- Metrics bar directly below name — key numbers immediately visible
- Metric-led bullets — every key line front-loads a number or outcome
- F-pattern optimized — designed around how hiring managers actually read resumes
Automated Visual QA Review
The PDF is rendered to an image and analyzed by AI vision to catch layout and formatting issues before delivery.
- Checks for orphaned text (single words isolated on their own line)
- Verifies column balance and white space distribution
- Flags crowded or awkwardly wrapped sections
- Confirms header fits on a single line
If issues are found → fix JSON data → re-render → re-QA. Repeats until clean.
Resume Emailed with Full Summary
The final PDF is emailed directly to the user along with a complete tailoring summary.
- PDF attached to email
- Match score with breakdown of matched vs. gap keywords
- Summary of key tailoring decisions made
- Interview recommendations based on role requirements
Final Outputs
output/YYYY-MM-DD-company-title.pdfIncluded in delivery emailjobs/YYYY-MM-DD-company-title.mdAI-Assisted · No proprietary or personal data · williamcastro.dev