Privacy Policy

The data controller is JobsEden Technologies Limited ("JobsEden"), which operates VerifiedTalents.ng from Nigeria.

• General contact: [email protected]
• Data protection contact: [email protected]

Account data. Your name, email address, and either a password (stored only as a hash, never in plain text) or your Google sign-in identity. If you sign in with an email code, the code is stored only as a hash and expires within minutes.

CV data. The CV file you (or an employer, with your consent) upload, the text extracted from it, and the structured claims our extraction produces: full name, email, phone number, education history (institution, degree, course, graduation year, matriculation number), NYSC details (service year, certificate number, call-up number, state), skills, work history, and links such as GitHub or portfolio URLs.

Verification inputs. Details supplied to complete a check, such as a matriculation number, surname, institution name, date of birth, or NYSC certificate number.

Profile data. If you build a profile: your display name, a headline, a profile photograph, your phone number, your date of birth, the profile address (slug) you choose, and the badge code we generate for you. Which of these become public is covered in section 11.

Consent records. When consent is given, attested to, or revoked, we record who acted, when, and the IP address and browser details of the request. This is the evidence that a check was authorised.

Verification evidence. The answer each source returns, and a stored copy of the raw response behind it (for example the results page the NYSC portal serves), truncated to a size limit. A raw response is captured as the source sent it, so it can contain data about other people that happened to appear on the same page. It is kept as the evidence behind your report; see sections 6 and 7.

Organisation and invitation data. For team accounts: the organisation, its members, and the email address of anyone invited to join it. We store an invited address even when it belongs to someone who has no account with us, so the invitation can be accepted or declined; the audit log keeps a record of the invitation afterwards.

Payment records. Transaction references, amounts, and refund records from Paystack. We never see or store your card number.

Technical and security data. IP addresses and browser details used for rate limiting and abuse prevention, and an append-only audit log of every significant action in the system.

• Consent (NDPA): running verification checks on a CV. Self-uploads record your own consent; employer uploads require a recorded attestation of yours. You can withdraw at any time (section 8).
• Contract: creating and operating your account, taking payment, delivering reports and refunds.
• Legitimate interest: fraud prevention, rate limiting, security monitoring, and the audit trail that protects both candidates and employers from tampered results.
• Legitimate interest (team invitations): when an account holder invites a colleague to their organisation, we process the invited email address so the invitation can be delivered, accepted, or declined, and so we can show who invited whom. If you were invited and want the address removed, write to us and we will delete the pending invitation.
• Legal obligation: where Nigerian law requires us to keep or disclose records.

We share data only to run the service. The recipients, and exactly what each one receives, are:

• OpenAI (AI processing): the text of the uploaded CV, sent to extract its claims. For paid skills checks, the candidate's name, claimed skills, and the public links on the CV are also processed by an OpenAI model.
• Cloudflare (file storage): the CV file itself, stored in a private R2 bucket under a randomised name. There is no public URL to a CV, ever, and the storage location is never exposed through our API.
• Resend (email delivery): your email address and the content of the email. This version of the service sends two kinds of email: your one-time sign-in code, and, when an employer uploads a CV that is not their own, the notice to the address on that CV with its consent-withdrawal link (section 10).
• Paystack (payments): your email and transaction details when you pay.
• Google (sign-in): your Google account identity, only if you choose Google sign-in.
• NYSC portals (verification sources, in Nigeria): for a senate-list check we send the institution, matriculation number, surname, and, where supplied, date of birth, exactly as a person checking manually would.
• GitHub and sites linked on the CV: for paid skills checks we fetch only public pages at URLs that appear on the CV (or share their origin) and query the public GitHub API. We never log in as you or send credentials anywhere.
• The other party to a verification: if an employer initiates a check with your consent, that employer sees the report.

Some of the providers above process data outside Nigeria: OpenAI (AI extraction), Cloudflare (file storage), Resend (email), Google (sign-in), and parts of Paystack's payment infrastructure. Under the NDPA this is a restricted transfer, and we want to be plain about it: when you upload a CV, its contents are processed on servers outside Nigeria for the purposes described in section 4 and no other.

We limit each transfer to what the recipient needs (the CV text goes to OpenAI; the file bytes go to Cloudflare; neither gets the other), and we rely on the recipients' contractual data protection commitments. Formal NDPA transfer safeguards (approved mechanisms and, where volume warrants, registration with the Nigeria Data Protection Commission) are a commitment we are actively working toward, not yet a completed certification, and we say so here rather than overclaim.

These measures are implemented in the software today:

• Passwords and one-time email codes are stored only as hashes and compared in constant time.
• API keys are stored only as SHA-256 hashes; the full key is shown once and never kept.
• CV files live in private object storage under randomised names. No public URL exists and the storage reference never appears in any API response.
• Consent and revocation tokens are cryptographically random and act as the sole credential for revocation.
• An append-only audit log records every significant state change; application code can only add to it, never edit or delete.
• Verification evidence is snapshotted for dispute review. The summary and evidence note we show you have third-party email addresses stripped out before storage.
• Rate limiting on sign-in, code issuing, and code verification bounds guessing and abuse.
• Fetches of CV-linked URLs are guarded against reaching private networks and never carry authentication.

Short-lived security data is deleted by the software on these horizons, measured from when the record was created:

• One-time email codes (with the requesting IP address): about eight hours.
• Sign-in and registration rate-limit records (IP addresses): about eight hours.
• Google sign-in state and session exchange records: about eighty minutes.
• Internal rate-limit keys that embed an IP or email: about one day.

A scheduled sweep now runs those deletions on a timer inside the service, rather than only on the back of the next sign-in or code request. The horizons above are what the software applies; a sweep that is interrupted is picked up by the next one.

Longer-lived data:

• CV files, extraction results, and check results are kept while your account and the reports that depend on them exist.
• Raw evidence snapshots — the copy of what a source returned — are kept for 60 days after the check completes and are then deleted by the scheduled sweep. That is the dispute window: after it, the report keeps the sanitised outcome (what was checked, against which institution, and what came back) but not the raw source page, which can contain other people's details.
• Consent records and the audit log are kept for the life of the records they protect, so that who authorised what can always be established. When you delete a verification, its consent records go with it and the IP address and device string the audit log had copied from them are erased; the log still shows the account, the action, and the time.
• Payment and refund records are kept as long as financial regulations require.

Automatic time-based deletion of old CVs and reports is not yet implemented. Deletion happens when you ask for it: you can delete any verification yourself, from its page on your dashboard, at any time (section 8). We state that plainly rather than promise a schedule the software does not enforce.

You have the right to:

• Access the personal data we hold about you and get a copy.
• Correct data that is inaccurate or incomplete.
• Delete your data. Deleting a verification is self-service, from the verification's own page — on your dashboard for a CV you uploaded about yourself, and on the employer dashboard for a CV you uploaded about someone else. It removes the CV file from storage, the extracted text and claims, every check result with its evidence, and any supplied details, and it retracts anything published from it to your public profile or badge, immediately. Payment and refund records are kept as financial records. The audit trail is kept as an integrity record: it keeps which account did what and when, so it can always be established who authorised a check — but the IP address and device recorded when consent was given or withdrawn are erased along with everything else. One honest limit: while a refund is still being retried automatically, deletion waits, so that deleting can never make money owed to you disappear. If those automatic retries are exhausted and the refund has to be settled by hand, deletion goes ahead and the amount stays recorded against the payment — your right to erasure is never held hostage to our payment provider.
• Restrict or object to processing.
• Receive your data in a portable format.
• Withdraw consent at any time. A self-upload consent can be withdrawn from your account. If someone else uploaded your CV, we email you a withdrawal link that works with no account at all — and if we could not reach you, the data protection address below withdraws it for you. See section 10. Withdrawal stops future checks; it does not undo checks already lawfully run.
• Not be subject to a decision based solely on automated processing that significantly affects you, and to be told when automated processing is being carried out on your data. Section 9 explains what is automated here and what is not.
• Complain to the Nigeria Data Protection Commission (NDPC) if you believe we have mishandled your data.

To exercise any of these, email [email protected] from the address on your account, or from any address with enough detail for us to verify it is you. We aim to respond within 30 days. Deleting a verification is self-service from your dashboard (described above); access, rectification, and whole-account erasure are currently handled by our team on request rather than by a self-service button. The right is the same either way.

Two parts of this service are automated, and you should know about both:

• Extraction. An OpenAI model reads the text of your CV and turns it into structured claims (institutions, degrees, matriculation numbers, NYSC details, skills, employers). No human reads your CV to produce that output.
• Verdicts. Software compares what a source returned against the claim and assigns the outcome — verified, not found, needs review, unverifiable, or self-reported. That assignment is automatic; no person reviews it before it appears on your report.

We do not make any decision about you. We do not score, rank, rate or profile candidates, and we produce no hiring recommendation. The report states what a source said at a point in time. Any decision — to interview, to hire, to reject — is made by the employer, on their own judgement and their own responsibility, and our Terms say so to them directly.

The software is built to fail towards "needs review" rather than a false negative: when a source is unreachable, shows a captcha, or changes its layout, the check says so instead of reporting "not found". A wrong "not found" can cost someone a job, and we treat that as the failure mode to design against.

An employer or recruiter can only run checks on your CV after attesting, on the record, that they have your consent. That attestation is stored with their identity, the time, and their IP and browser details, and the resulting report is flagged as employer-initiated.

A revocation credential is created for that consent and is never shared with the organisation that uploaded your CV. Presenting it back to us withdraws the consent immediately, with no account needed, and no further checks run on that CV.

That credential is now sent to you automatically. When an employer uploads a CV that is not their own, we email the address on the CV before any check can be paid for or run: who asked, what may be checked, and a link that withdraws your consent on the spot, with no account and no sign-in. Following it stops every further check on that CV and retracts anything already published from it.

So the manual route stays open, for that case and for any other. Email [email protected] — from any address, with no account — saying that your CV was uploaded and by whom, if you know. We will withdraw the consent and confirm back to you.

Publishing a profile is your choice; nothing is public by default. A published profile shows your display name, your headline if you wrote one, your profile photograph if you uploaded one, and your claim outcomes, such as "NYSC: verified". A profile photograph is personal data and it is published to the open internet, so treat uploading one as a deliberate choice; you can remove it or unpublish the profile at any time.

A published profile never shows your CV file, your email address, your phone number, your date of birth, or any document. Badge verification returns claim outcomes only, with no name, photo, or contact details of any kind.

If a personal data breach occurs that is likely to risk your rights, we commit to notifying the Nigeria Data Protection Commission within 72 hours of becoming aware of it, and to telling affected people directly and promptly, in plain language, with what happened and what we are doing about it.

The service is for people aged 18 and over. We do not knowingly process children's data; if you believe we have, contact us and we will delete it.

If we change this policy materially we will give notice on the site or by email before the change takes effect. Questions: [email protected], or see the contact page.