This policy covers what we collect, why we collect it, how we use AI on your data, who we share it with, how long we keep it, and the rights you have. We've written it to actually be readable — if anything is unclear, email us.
JobSphere ("JobSphere", "we", "us") is an AI-native job-seeking platform operated by the JobSphere project. Contact details are at the bottom of this page and on our About page. For the purposes of the EU GDPR and the UK GDPR, JobSphere is the data controller for personal data you provide when using the service.
We collect personal data in three ways:
If you sign in with Google, GitHub, or LinkedIn we receive your email, name, profile picture, and the provider ID. We do not receive your provider password.
Under GDPR, our processing rests on the following legal bases:
Contract
Providing you the service you signed up for — running your profile, storing your résumé, delivering applications to employers, computing fit scores.
Legitimate interest
Operating the service securely (fraud detection, rate limiting, server logs) and improving it (aggregate anonymous analytics).
Consent
For optional AI features like drafting cover letters from your résumé. You can opt out by not using the feature.
Legal obligation
Responding to lawful requests from authorities, tax/record-keeping, and safety obligations.
This section is the part most privacy policies gloss over. We don't. Whenever we send your data to a third-party model, we do it for a specific task, with the minimum input needed, and never for training.
You upload a résumé
Extracted plain-text of your résumé is sent to Claude (parse) + OpenAI (embedding).
An employer posts a job
Pasted URL content or raw text is sent to Claude (parse) + OpenAI (embedding) after our prompt-injection scanner sanitizes it.
You click "Draft cover letter"
Your name, headline, skills, the job title, company, and the job description are sent to Claude. Streamed back to you. Not saved until you hit submit.
You submit an application
Your résumé embedding + the job embedding are compared locally. A short rationale is generated by Claude. No raw résumé text is re-sent.
You view recommendations
Only your stored profile embedding is used to query MongoDB Atlas Vector Search. No prompt is sent.
Prompt-injection protection. Every job posting is scanned before any LLM sees it. If we detect patterns like "ignore previous instructions", role hijacks, or fit-score manipulation attempts, the posting is flagged and the detected patterns are stripped before forwarding the remainder to the model. This protects your AI-generated artifacts (cover letters, fit scores) from being influenced by a malicious poster.
We share personal data only in the following cases:
| Provider | Purpose | Data shared |
|---|---|---|
| MongoDB Atlas | Primary database + Vector Search | All profile, résumé, job, and application data |
| Vercel | Hosting, edge functions, cron, analytics | Request metadata, rendered HTML, aggregated usage |
| Anthropic | Claude LLM (chat, parsing, drafting) | Résumé text, job text, profile snippets — per task |
| OpenAI | Embeddings for vector search | Résumé text, job text — per task |
| Google / GitHub / LinkedIn | OAuth sign-in | Your email, name, profile picture (only if you use that provider) |
When you apply to a job, the employer who posted that job sees your application data — name, contact info, résumé, cover letter, eligibility answers, screening responses, and your fit score breakdown. They don't see applications you've sent to other employers. They don't see your profile until you apply (unless you've set your profile visibility to Public).
We may disclose data if required by law, court order, or to protect rights and safety. We will notify you unless legally prohibited.
Under GDPR / UK GDPR / CCPA / Quebec Law 25, you have the rights below. Most are exercised directly from your profile; for anything we can't do with a button, email us.
Access + portability
Request a machine-readable export of your data.
Correction
Edit any field on your profile yourself, or ask us to correct it.
Erasure
Delete your account and all associated data from your Profile → Account → Danger zone.
Restriction / objection
Ask us to pause specific processing (e.g., LLM drafts) while we review concerns.
Withdraw consent
Turn off AI features any time; prior AI-generated artifacts remain in your history unless you delete them.
Complaint to authority
Lodge a complaint with your local data-protection authority (ICO, CNIL, OPC, OAIC, etc.).
We use the bare minimum:
We do not use third-party advertising cookies, tracking pixels, or cross-site fingerprinting.
When your data moves outside your home country — for example, to Anthropic's or OpenAI's US-based APIs — we rely on Standard Contractual Clauses (SCCs) and the providers' own adequacy commitments to protect the transfer. MongoDB Atlas lets us pin your primary region at the cluster level.
JobSphere is not intended for users under 16. We don't knowingly collect data from children. If you believe a minor has created an account, email us and we'll remove it promptly.
We'll post any material changes here and update the "last updated" date at the top. For significant changes that affect how we use your data, we'll notify signed-in users in the app at least 30 days before they take effect.
Email us directly. We read everything and reply within 30 days (usually faster).
vishalpopat8@gmail.comThis policy was last updated on April 18, 2026.