Privacy Policy.

Effective: July 5, 2026

The short version

  • We never sell your personal data. No ad networks, no data brokers. Our product-analytics tool is PostHog.
  • As a guest we count anonymous pageviews and button clicks so we can measure landing-page traffic and referrers — with no session replay and no Person profile until you sign up. Once you sign up, we track product usage (page views, saves, applies, searches, and a masked session replay) to improve the product.
  • We only receive resume data when you upload a resume. We then keep the file, its full extracted text, an AI-parsed version, and your prior uploads.
  • Your resume text is sent to Anthropic (our AI provider) to power parsing, tailoring, and cover letters. Anthropic does not train on API data by default.
  • Deleting your account immediately erases your identity details and disables sign-in, but some records (resumes, AI outputs, activity history) are currently retained afterwards — section 12 explains exactly what happens.
  • Questions or requests? Email hello@neurohire.ai.

1. Who we are and what this policy covers

In plain English: NeuroHire runs this site; here's who's responsible for your data and how to reach us.

NeuroHire (“NeuroHire”, “we”, “us”) operates the job-search platform at neurohire.ai. The data controller for the personal information described in this policy is NeuroHire, the operator of neurohire.ai.

This policy covers the neurohire.ai website and app, including public job browsing without an account. Some of the collection described below — analytics and browser storage in particular — starts before you sign up, so this policy applies to visitors as well as account holders.

You can reach us at hello@neurohire.ai for any privacy question or request in this policy.

2. Information you provide

In plain English: Your name, email, profile, and resumes — including the text and AI-parsed versions of every resume you upload.

Registration. Full name, email address, and password. Your password is stored only as a bcrypt hash — never in plain text.

Profile. Phone number, skills (free-text tags), experience level, preferred locations, preferred job types, and an avatar image URL. All of these are optional and editable at any time.

Resumes. You can upload a resume as a PDF or DOCX file up to 10 MB (we validate the file type). For each upload we store: the file itself, the full extracted plain text, an AI-parsed structured version (work history, education, skills, contact information), and a file hash.

Important: When you upload a new resume, we keep the prior versions too — resume revision history is retained, including the extracted text and parsed content of earlier uploads. See section 12 for how this interacts with deletion.

AI outputs you generate. Tailored resumes (full section content, tailoring strategy, and ATS-score snapshots) and cover letters (full body plus the AI's reasoning) are stored per user and job. Draft tailoring responses are also cached for 7 days to make regeneration fast.

Profile autofill. After parsing a resume, we automatically fill empty profile fields from it — phone number, up to 50 skills, and experience level. You can edit or remove anything it fills in.

Waitlist. If you join the Resume AI waitlist we store your email address and where you signed up from. There are two ways in: typing your email on the landing page, or clicking the in-app “Get it first” pill — which submits your account email with one click. The only purpose is a single launch-announcement email (“No spam — one email when it ships”).

Important: The waitlist currently has no self-serve unsubscribe and no automated deletion path, and entries are stored indefinitely. If you want your email removed from the waitlist, email hello@neurohire.ai and we will remove it manually.

Support. If you email hello@neurohire.ai, we receive your message and email address. Our mailboxes are handled via Google Workspace.

3. Information collected from guests (no account)

In plain English: As a guest we count anonymous pageviews and button clicks to measure landing traffic. Session replay is off until you sign up.

When you browse without an account, PostHog (see section 7) captures anonymous pageviews and autocaptured interactions (button clicks, form submits, and link clicks) so we can measure landing-page traffic, referrers, country breakdowns, and conversion to signup. Search terms typed on the jobs board and pages you visit are recorded as event properties on those pageviews.

As a guest you do not have a Person profile in PostHog and no session replay is recorded. PostHog assigns a distinct identifier to your browser (stored as a ph_* cookie / localStorage entry) so repeat visits from the same browser aren't double-counted as unique. If you sign up later, the events from your pre-signup session are attributed to your new account.

When you submit the landing waitlist form we also record the join with the email you provided (source: landing).

Your browser also keeps a local activity store: jobs you viewed or dismissed and your last 10 searches. This lives on your device, not on our servers (section 5).

Important: Guest-to-account linking: if you try to save a job or confirm an application while signed out, that intent is stashed in your browser's session storage and is automatically written to the account you subsequently create or log into. Activity from before you signed up may therefore be linked to your account.

4. Information collected automatically

In plain English: We log IPs, devices, and request patterns to stop bots and break-ins — most of it auto-deletes within 7–90 days.

We process the following for security and fraud prevention (our legitimate interest — see section 10):

  • IP address and user-agent — captured at signup, login, and token refresh; held in your session record for the session's life (at most 7 days) and in security audit logs.
  • Device fingerprint — your browser may supply a fingerprint that we bind to your session; a mismatch invalidates the session. We also compute server-side request fingerprints (a hash of browser headers), retained for 30 days and indexed by IP and user ID.
  • Bot protection — every request's headers are analyzed; a short per-IP request history is kept for about 10 minutes; requests scoring above confidence thresholds are automatically blocked.
  • IP reputation — per-IP violation scores drive escalating automatic blocks (1 hour, 6 hours, 24 hours, or 7 days). Blacklisted IPs are stored indefinitely, and hidden honeypot endpoints permanently blacklist IPs that access them. If you believe you were blocked in error, email hello@neurohire.ai and a human will review it.
  • Brute-force protection — failed login attempts (including the email address attempted) are logged; 5 failures lock the account for 15 minutes, and per-IP locks apply too.
  • Security audit log — kept in Redis only: a global stream capped at 100,000 events, a per-user index kept 90 days, a per-IP index kept 30 days, and alerts kept 7 days. Staff with the admin flag can view these logs as part of internal security operations.
  • AI telemetry — every AI call logs the model used, token counts, latency, cost, and validation results tied to your user ID. Prompt content is not logged here.

Job activity on our servers (signed-in users). We store the jobs you save, the jobs you view — explicitly including “clicked Apply but didn't finish” — and the applications you tell us about, with their status over time.

Important: Application tracking is self-reported. When you click Apply, we open the employer's site and later ask “Did you apply?” — your dashboard statuses are records you maintain, not confirmations from employers. We also record that you clicked Apply even if you never confirm.

5. Information stored only on your device

In plain English: Some data (like preferences and certain profile answers) lives only in your browser and never reaches our servers.

The following never leaves your device — we cannot see it:

  • Job activity store (neurohire-user-store): saved, applied, viewed, and dismissed jobs (the last 500) plus your last 10 search queries. Persists until you log out or switch accounts.
  • Preference and UX flags: theme, board density, first-run tour completion, promo-pill dismissal, waitlist-joined flag, and nh_store_owner — your user ID, used to detect an account switch on a shared browser. This last key is not removed on logout.
  • Flag-gated profile fields (currently hidden in the UI): headline, location, about, experience and education entries, preferred title, minimum salary, relocation and workplace preferences — and notably work-authorization status and security clearance.
  • Session storage (cleared when the browser session ends): pending apply/save intents and banner dismissals.
Important: Device-only fields — including work-authorization status and security clearance — are never sent to our servers, but they are also not deleted by logging out or deleting your account. If you use a shared computer, clear the site's data in your browser settings to remove them.

6. Cookies and local storage

In plain English: Three login cookies we need, PostHog cookies for signed-in users we don't strictly need, plus browser storage — full table with lifetimes.
NameTypePurposeDuration
access_tokenCookie, HttpOnlyAuthentication~15 min
refresh_tokenCookie, HttpOnlyKeep you signed in7 days
csrf_tokenCookie, JS-readableCSRF protection; also used to detect signed-in state7 days
ph_*Cookie + localStorage (PostHog)Product analytics (signed-in users only; also used pre-signup on the landing waitlist form, /signup, and /forgot-password)~1 year
neurohire-user-storelocalStorageJob activity + recent searchesuntil logout
nh_store_ownerlocalStorageDetect account switch on shared browserindefinite
theme, nh_density, nh_tour_v1, nh_tease_seen, nh_waitlist_joinedlocalStoragePreferences/UI stateindefinite
nh_profile_* / nh_work_auth / nh_clearance etc.localStorageDevice-only profile fields (feature-flagged)indefinite
nh_pending_apply / nh_pending_track / nh_pending_save / nh_auth_from / banner flagssessionStorageApply-flow continuity, UI statebrowser session

The first three cookies are strictly necessary for the service to work and require no consent. The PostHog cookies are not strictly necessary — see section 7 for how product analytics currently operates.

7. Analytics (PostHog)

In plain English: PostHog sees what you search and which jobs you view, save, and apply to — plus a masked video of your session. Only after you sign in. Your name and email are attached to your Person profile (that's how we know it's you), but input fields (passwords, notes) are masked in replays.

We use PostHog, a product-analytics platform provided by PostHog, Inc. Data is stored on PostHog's US Cloud infrastructure. PostHog is certified under the EU-US Data Privacy Framework for transfers from the EU, UK, and Switzerland.

When PostHog loads: only after you sign in (or, in three narrow pre-signup cases described in section 3, when you submit the landing waitlist form or open the sign-up / forgot-password pages). If you never sign in and stay off those three pages, PostHog never loads and no product analytics are recorded.

To be transparent about granularity, PostHog receives event-level data, not just aggregates:

  • Page URLs, including query strings.
  • Raw free-text search terms and result counts.
  • Per-job view, save, apply, and dismiss events with the job ID, job title, and employer name.
  • Account lifecycle events: sign-up, login, logout, email verification, profile update, account deletion, and password-reset request (event names only — no profile content).
  • “Not interested” relevance signals, which we may use to improve ranking and recommendations.
  • Person properties attached to your PostHog identity: email, first and last name, verification status, sign-up date, and running counts of jobs you have saved, applied to, viewed, or dismissed.
  • Session replay: a masked video of your interactions with the site. All input fields (email, password, search, resume text, notes) are obscured. Retained by PostHog for approximately 30 days.
Important: We do not currently show a cookie-consent banner. Our position is that guests are not tracked at all (PostHog never loads), and signed-in users have accepted our Terms and this Privacy Policy — which explicitly describe PostHog — as a condition of creating an account. You can still block PostHog with browser settings, network filters, or extensions.

Data sent to PostHog may be processed on PostHog's US infrastructure outside your country. The transfer relies on PostHog's certification under the EU-US Data Privacy Framework for EU/UK/Swiss data. To request deletion of your PostHog data, email hello@neurohire.ai.

8. AI processing

In plain English: Your resume text goes to Anthropic to power parsing and tailoring; job listings are enriched by AI that can get things wrong.

Anthropic PBC processes resume data on our behalf: the full text of your resume (and, for hard-to-read PDFs, the raw PDF file) is sent to Anthropic's API for parsing, and your parsed resume plus the job description are sent when you generate a tailored resume or cover letter. Anthropic does not use data submitted through its API to train its models by default. Retention of API inputs is governed by our agreement with Anthropic.

Job-catalog AI (no user data). Separately, our job catalog is enriched by AI that never receives your personal data: Google Gemini extracts structured fields (salary, visa sponsorship, clearance, qualifications) from scraped public job postings, and OpenAI computes job-description embeddings for search. These systems process job content only.

Profiling and automated decisions. Match scores are computed algorithmically from rule-based signals (skills, experience, and location overlap). They are informational only — they are not an employment decision and have no legal or similarly significant effect on you. Relevance feedback you give (like “not a fit”) may inform future ranking. The only fully automated decisions we make are security blocks (section 4); you can contest any block by emailing hello@neurohire.ai.

9. Processors and recipients

In plain English: Anthropic, Google, and our hosting provider — no ad networks, no data brokers, no selling.
  • Anthropic PBC — AI resume parsing, tailoring, and cover letters (receives resume content, which is personal data).
  • PostHog, Inc. — product analytics and session replay (receives behavioral data and Person properties for signed-in users, per section 7). Data stored on PostHog's US Cloud.
  • Google LLC — Google Workspace Gmail SMTP for transactional email delivery (receives your email address and message content).
  • Contabo — hosting provider. Our databases, cache, and search index run on Contabo infrastructure, and resume files are stored on the server's filesystem. Contabo operates data centers in Europe and the United States.

That is the whole list. We use no ad-tech, no fingerprinting SDKs, and no data brokers. Fonts are self-hosted (no runtime requests to Google Fonts). Session replay is provided by PostHog for signed-in users only, with all input fields masked. We do not sell or rent your personal information.

10. Purposes and lawful bases

In plain English: Every use of your data mapped to a legal reason.
PurposeLawful basis (GDPR)
Providing your account, profile, job saving, and application trackingPerformance of a contract
AI resume features you initiate (parsing, tailoring, cover letters)Performance of a contract
Transactional email (verification, password reset)Performance of a contract
Security and fraud prevention (fingerprinting, bot and IP-reputation systems, brute-force protection, audit logs)Legitimate interests
Improving ranking and recommendations from feedback signalsLegitimate interests
Product analytics and session replay (PostHog, signed-in users only)Legitimate interests, with a specific consent step folded into signup (accepting Terms + this Privacy Policy).
Waitlist (one launch email)Consent

11. How long we keep data

In plain English: Exactly how long each kind of data lives.
DataRetention
Access token~15 minutes
Refresh session (including its IP address and user-agent)Up to 7 days
Email verification link24 hours
Password-reset link1 hour
Draft tailored-resume cache7 days
Security alerts7 days
Per-IP audit index, request fingerprints, IP reputation scores30 days
Per-user audit index90 days
Important: Some records currently have no fixed retention limit and no automated purge: resume revisions (including raw text and parsed content), tailored resumes, cover letters, application/saved/viewed job records, AI telemetry, waitlist emails, and IP blacklist entries. We plan to introduce retention jobs; until then, these are retained indefinitely unless you ask us to erase them (section 12).

12. Your rights and how deletion works

In plain English: Delete your account from your Profile page — here's precisely what gets erased and what doesn't.

Depending on where you live, you may have the right to:

  • Access — get a copy of your data.
  • Rectification — fix incorrect data (you can edit your profile directly).
  • Deletion — delete your account and personal data.
  • Portability — receive your data in a machine-readable format.
  • Restriction and objection — limit certain processing.
  • Withdraw consent for consent-based processing at any time.
  • Complain to your data-protection supervisory authority.

We respond to rights requests within the timeframes required by applicable law.

Where to delete your account: on your Profile page (account section), not in Settings.

What account deletion does today. Deleting your account immediately scrubs the identity fields from your user record (email, name, phone, avatar, skills, preferences), deactivates login, and revokes your sessions.

Important: Account deletion currently does not erase everything. Resume files and parsed resume content, tailored resumes, cover letters, and your job-activity history are retained after deletion, and a waitlist entry (if you joined) is not removed. Likewise, “Delete resume” removes only the latest revision — older revisions remain. To request full erasure of these records, email hello@neurohire.ai and we will process it manually.

Export. You can request a machine-readable export of your data by emailing hello@neurohire.ai. Our automated export covers your profile, applications, saved and viewed jobs, resume metadata, and session records (with IP and user-agent); tailored resumes, cover letters, parsed resume content, waitlist entries, and audit logs are not yet included in it — ask and we will provide those separately.

Residual device data. Server-side deletion does not clear your browser's local storage (section 5). On a shared device, also clear the site's data in your browser.

13. California privacy rights (CCPA/CPRA)

In plain English: Your CCPA rights and our answers.

For California residents, the categories of personal information we collect (as defined in Cal. Civ. Code §1798.140) are: identifiers (name, email, phone, IP address), professional or employment-related information (resume content, work history, applications), internet or network activity (job views, searches, usage data), and inferences (match scores, relevance signals). We collect them from you directly and automatically from your use of the service, for the business purposes described in sections 2–10.

We do not sell personal information. Our analytics provider, PostHog, processes data only on our behalf as a service provider — we use no ad networks and no cross-context behavioral advertising. We do not currently process Global Privacy Control (GPC) signals.

You may exercise the rights in section 12 without discrimination — we will never deny you the service, charge different prices, or degrade quality because you exercised a privacy right. You may also submit a request through an authorized agent by emailing support@neurohire.ai; we will verify the request with you directly before acting on it.

14. International data transfers

In plain English: Where your data is stored and how cross-border transfers are protected.

Our infrastructure is hosted by Contabo, which operates data centers in Europe and the United States. Our processors Anthropic and Google may process data in the United States and other countries. Where personal data crosses borders, we rely on the safeguards available for each processor — such as standard contractual clauses or the EU-US Data Privacy Framework (PostHog's certification is described in section 7).

15. Security

In plain English: Hashed passwords, locked-down cookies, short-lived sessions, automated abuse defenses.
  • Passwords are hashed with bcrypt and never stored in plain text.
  • Authentication uses HttpOnly cookies, short-lived access tokens (~15 minutes) with rotating refresh tokens, and immediate session revocation. No credentials are kept in browser web storage.
  • Double-submit CSRF protection on all state-changing requests.
  • Security headers, including X-Frame-Options DENY and a Permissions-Policy that disables geolocation, camera, and microphone.
  • Email is delivered over TLS (STARTTLS).
  • Automated bot, brute-force, and IP-reputation defenses (section 4).
  • Access to security data is restricted to admin staff.

To be candid about limits: resume files are stored on our server's filesystem protected by access controls, but they are not currently encrypted at rest, and we do not currently run automated vulnerability scanning. We are stating this plainly rather than overclaiming.

16. Data breach notification

In plain English: If your data is breached, we'll tell you.

If a security breach affects your personal data, we will notify you and the relevant regulators without undue delay, as required by applicable law.

17. Children

In plain English: You must be 16+.

NeuroHire is not directed to individuals under 16, and you must be at least 16 to use the service. We do not knowingly collect personal data from children; if we learn we have collected data from a child in error, we will delete it.

18. Changes to this policy and contact

In plain English: We'll date and announce policy changes.

We will update this policy as the service evolves. For material changes, we will give notice via the service (such as a banner or in-app notice) and update the effective date at the top of this page, which always reflects the current version.

For any privacy question or request: hello@neurohire.ai.