Resume and CV Data Processing Notice
Detailed explanation of deterministic parsing, OCR, optional AI enrichment, and candidate suggestions.
This policy draft is tailored to AI Tathastu's verified architecture and engineering practices. Under India's phased legal framework (Digital Personal Data Protection Act, 2023 and DPDP Rules, 2025), substantive obligations commence in scheduled tranches.
Any item labelled with an [Awaiting Verification] tag designates operational data (such as formal legal entity name, designated grievance officer contacts, or vendor contractual tiers) reserved for verified business confirmation. This platform makes no unverified certification claims.
Overview#
Applies when you upload a PDF or DOCX on the AI Tathastu onboarding or profile flow.
Before you upload#
Review the document for unnecessary information about yourself or other people. Do not upload confidential employer records, government identity numbers or sensitive data unrelated to your educational profile. You can continue onboarding manually without a CV.
Permission A — required to use resume upload#
> [ ] I allow AI Tathastu to upload, store and process my CV to prepare reviewable profile suggestions.
We may use PDF/DOCX parsers, OCR, layout analysis, rules, dictionaries, URL detection and document validation. The file and intermediate records may be stored before you review the resulting suggestions. This permission does not authorize external generative-AI enrichment or public profile fetching.
Permission B — optional external AI assistance#
> [ ] I separately allow eligible external AI providers to process relevant CV text to help interpret unclear sections.
When enabled, the service may send selected text or structured context to the provider named in the current disclosure. This option is not required for baseline deterministic extraction. You may decline or withdraw permission for future optional AI calls; a prior completed call cannot be undone. If a provider is ineligible for your age or has unsuitable data-use terms, the service must not use it even when a checkbox is checked.
Information that may be suggested#
Contact details, university/school, degree, academic dates/grades, technical skills, projects, employment/internships, leadership, certifications, achievements, research, patents/books and profile URLs where supported by the document. Not every category will be extracted. Extracted values may be incorrect, incomplete or assigned to the wrong record.
Review and verification#
We display suggestions for acceptance, correction or rejection. A document match is not proof that a claim is true or that a platform account belongs to you. User-confirmed and independently verified items have distinct labels. Some draft/pending-suggestion state may already exist before final review; it must not be treated as verified profile information.
Third-party infrastructure#
Application processing may occur on Vercel and Render; file/metadata storage may occur in configured object storage and Neon PostgreSQL. Actual processor regions and retention times are stated in verified registers linked from the Privacy Policy. Optional model providers are a separate recipient category and are not guaranteed to have identical training or retention policies.
Detected digital links#
GitHub, LinkedIn, Kaggle, research and other links may be suggested from text or embedded hyperlinks without fetching their profiles. Retrieval, connecting and importing are separate activities governed by the Digital Sources Notice.
Deleting or replacing your CV#
Use available deletion controls or Privacy Requests. Deleting a file is not the same as retracting information you separately confirmed into your profile. The verified deletion schedule explains retention of document copies, backups, security records and reviewer evidence. We do not promise instant deletion from every system.