Transparency & Limitations

How CAIJA works and where its boundaries are

We believe you deserve to understand how our AI makes decisions about your career materials. This page explains what CAIJA does, how it works, and — just as importantly — what it cannot do.

How CAIJA Analyzes Your Resume

CAIJA uses a multi-stage AI pipeline to evaluate the fit between your resume and a target job description:

  1. Text extraction — We parse your resume and the job description into clean, structured text.
  2. Requirement classification — The job description is broken into categories: hard skills, soft skills, experience levels, certifications, domain knowledge, and more.
  3. Evidence extraction — Your resume is scanned for concrete, verifiable evidence that supports each requirement.
  4. Matching & scoring — Each requirement is matched against your evidence. Scores are computed based on match strength, requirement importance, and gap severity.
  5. Recommendations — Tailoring suggestions are generated based on identified gaps and strengths.

What the Scores Mean

85+ — Strong Match

Your resume demonstrates strong alignment with this role. Consider applying.

80–84 — Moderate Match

Your resume has potential but could benefit from targeted tailoring before applying.

Below 80 — Weak Match

Significant gaps exist between your resume and this role's requirements. Review the gap analysis before deciding.

Our Truthfulness Commitment

Evidence-based only. Every claim in your report is tied to specific text found in your resume. We never infer qualifications you haven't stated.

No fabrication. Recommendations suggest how to truthfully reframe your experience — never how to fabricate credentials or experience you don't have.

Gaps are real. If we identify a gap, it means we could not find supporting evidence in your resume text. The gap might exist because the experience isn't on your resume, or because it's described in terms the AI didn't recognize.

Known Limitations

CAIJA is a powerful tool, but it has boundaries you should understand:

  • 1.
    Text only. CAIJA analyzes the text content of your resume. It cannot evaluate formatting, visual design, layout quality, or how your resume looks to a human reader.
  • 2.
    Not a real ATS. CAIJA simulates how a resume might perform in screening, but it is not an actual Applicant Tracking System. Real ATS behavior varies by vendor and employer configuration.
  • 3.
    AI interpretation limits. The AI may not recognize industry-specific jargon, proprietary tools, or niche qualifications. If a gap seems incorrect, it may be a vocabulary mismatch rather than a true gap.
  • 4.
    Job description quality matters. Analysis quality depends on the completeness and clarity of the job description you provide. Vague or incomplete postings will produce less actionable results.
  • 5.
    No networking or soft factors. Hiring involves referrals, cultural fit, interview performance, and many other factors that resume analysis alone cannot capture.
  • 6.
    Point-in-time analysis. Results reflect the resume and job description provided at the time of analysis. Job postings change, and your resume should evolve with new experience.

AI Model Information

CAIJA uses OpenAI language models (GPT-5.4 family) for analysis. Model outputs are deterministic where possible (temperature=0) to ensure consistent scoring across repeated analyses of the same resume and job description. Results are cached to prevent score drift within the same analysis session.

Feedback

If you believe an analysis result is inaccurate or misleading, we want to know. Contact us at [email protected] with details about the issue so we can improve.