STORINarrative Intelligence
Exit

Positioning Report · Data & AI

How AI sees your Career Blueprint data

You share strengths, goals, and sometimes a resume. Here is exactly what we withhold from AI on resume upload, what still goes to models as you wrote or spoke it, and where to read the formal policies.

Promise

What we withhold on resume upload

When you upload a resume, we separate contact identity and pattern-mask it before resume-analysis, counselor, and “suggest fields” models see the file:

  • Name → APPLICANT
  • Email → [EMAIL]
  • Phone → [PHONE]
  • Major profile URLs (LinkedIn, GitHub, and similar) → [LINK]

Schools, employers, projects, metrics, and skills stay in the text — we need them to give positioning advice. Your real name and contact fields are kept in STORI's database for your report, PDF, and admin support — not re-injected into those resume LLM prompts.

Example

Before and after (fake sample)

As uploaded (illustrative)Priya Sharma
priya.sharma@email.com · +91 98XXX XXXXX
linkedin.com/in/priyasharma

Built a campus waitlist that hit 2,000 signups…
What resume AI typically seesAPPLICANT
[EMAIL] · [PHONE]
[LINK]

Built a campus waitlist that hit 2,000 signups…

Masking is best-effort pattern matching. If a name is missed on the page, it may still appear. Address and date of birth, when extracted, are stored for your session but are not fully stripped by today's masker.

Timeline

What leaves your device, and when

  1. Answers (Talk) — Your five written or spoken answers are processed as you wrote or spoke them so the report can quote your words. We do not run the same resume masker on those transcripts today.
  2. Voice (“Speak it”) — When you use the mic, audio is transcribed by our speech provider (OpenAI Whisper). Recording is not “local only until Build.”
  3. Resume (Context) — On upload, contact identifiers are separated and masked for resume-related AI passes as described above.
  4. Build — Report generation uses your answers, optional links, and (when present) masked resume text for counselor / fuel helpers. Public URLs you add — and any successfully scraped public page text — may be included in generation prompts unmasked.
  5. Unlock — Your named Positioning Report is assembled for you (and share/PDF), using identity stored separately where needed.

Honesty

What we never claim / what we still do

WE DO

  • Withhold name / email / phone / major profile links from resume AI when detection works
  • Keep contact fields in our database for your report and support
  • Tell you clearly when voice transcription is used

WE DON'T CLAIM

  • “All PII is removed before any AI”
  • Talk answers are sanitized the same way as resumes
  • LinkedIn / portfolio scrapes are anonymized
  • We train foundation models on your session (see Privacy Policy for processors & retention)

Formal policies

This page explains Career Blueprint / Positioning Report mechanics. The company-wide documents still apply:

Privacy Policy · Terms & Conditions

← Back to Focus

Last aligned with product behavior: August 2026. If engineering changes masking, this page updates in the same release.