AI Disclosure
How Truen Uses AI
Truen uses AI to turn selected engineering evidence into structured hiring context. These outputs are decision support, not automated hiring decisions.
Last updated: June 7, 2026
These pages are product disclosures and working policy drafts. They are not a substitute for legal advice and should be reviewed by counsel before public launch.
Where AI is used
- Truen uses AI to analyze selected GitHub repository evidence and produce a Truen Profile with narrative, capability dimensions, confidence, evidence references, weaknesses, flags, limitations, and interview follow-ups.
- Truen may use AI to generate job-relative qualification analysis for an employer after a student applies to a job.
- Truen does not currently infer AI collaboration style from GitHub history alone. The live product shows this section as not yet assessed until dedicated AI tool ingestion exists.
What AI receives
- AI prompts may include selected repo metadata, commit messages, pull request titles and bodies, issue titles and bodies, file paths, selected code samples, selected markdown documentation, and generated repo insights.
- The system is designed to prefer insufficient evidence over unsupported conclusions, but AI-generated text can still be incomplete or incorrect.
- Profile outputs may include selected code snippets or commit message samples as evidence when a student applies to a job.
Decision-support limits
- Truen does not make final hiring decisions and does not guarantee candidate performance, interviews, or employment outcomes.
- Employers must use human review and independent judgment before making hiring, rejection, or interview decisions.
- A qualification fit result is not a score, rank, or automatic rejection. It is a role-specific interpretation of available evidence.
Known limitations
- GitHub evidence can be incomplete, sparse, squashed, private, collaborative, or shaped by school, employer, open-source, or NDA constraints.
- Selected repositories may not represent the full candidate. Thin evidence should lower confidence rather than force a positive or negative conclusion.
- Qualification analysis may be regenerated by employers and can differ from the frozen profile snapshot unless the product later freezes fit inputs too.
Candidate review and disputes
Before broad launch, Truen should provide a clear way for candidates to report inaccurate profile claims, request correction, request deletion, or dispute outputs used in an application review.
Questions about these disclosures? Contact the Truen team before connecting repositories or applying to a role. You can also review the Privacy Policy, Terms, AI Disclosure, and Subprocessors.