Data-Usage Policy

Effective date: 2026-07-22

This policy is the plain-English companion to our Privacy Policy. It exists because Meridino (meridino.com), operated by Feld & Co LLC, handles two kinds of data that deserve a direct explanation rather than a paragraph buried in a general policy: the job postings we collect, and your resume.

It does not replace the Terms of Service or the Privacy Policy. Where this document and the Privacy Policy describe the same processing, they are meant to agree; if they conflict, the Privacy Policy controls.

1. Where our job data comes from

We collect job postings only from the public, first-party career-board endpoints operated on employers' behalf by their own applicant tracking system providers. These are the same feeds that power the "Careers" page on the employer's own website.

As of the effective date, the platforms we collect from are:

• Greenhouse
• Lever
• Ashby
• SmartRecruiters
• Workable
• BambooHR
• Teamtailor

We are building a Workday connector. Until it is live we do not collect Workday postings; Workday will be added to this list, with an updated effective date, when collection begins.

How we do it, precisely:

• We call each employer's own public, unauthenticated job-list endpoint on the platform above.
• We use no credentials, no API keys, no logins, and no paid access to obtain postings. If a platform's data requires an account or a key, we do not build an adapter for it.
• We do not scrape HTML career pages, and we do not operate a headless browser to work around a site's controls.
• We do not take postings from third-party job aggregators or job boards. We do not source from Indeed, LinkedIn, ZipRecruiter, Glassdoor, Monster, or any similar aggregator, and we hold no agreement, feed, or scraper for any of them.
• We rate-limit ourselves per platform, honor a platform's rate-limiting responses, and skip a board rather than hammer it.
• We periodically re-check whether a posting is still live and mark it dead when it is not, so that work you do is done against a real, current opening.

We also use a web-search API to find which employers to follow. To learn which companies publish on the boards above, we run search queries that we compose ourselves against a search provider (Brave Search, and Google Programmable Search as a secondary). Those queries are about job titles and board domains. No user data of any kind is sent to a search provider: not your resume, not your profile, not your identity, and not the text you type into Meridino's own search box. The search provider returns addresses, which we use only to identify a company's board; it never fetches or supplies posting content to us.

The postings pool is shared and is not personal data about you. It exists independently of any account. What is personal to you is the layer on top: which postings you were shown, which you saved, what you recorded in your tracker, and what our engine computed for you.

Jobs you add yourself are a separate, private category, never pooled. Your tracker lets you record a job you found anywhere, by pasting its link, typing its title and company, pasting its description text, or any combination. When we did not collect that posting ourselves from a first-party board above, we do not fetch the link you paste, we do not verify it, and we do not add it to the shared postings pool. It is stored as your own private note: visible only to you, never shown to any other candidate, never returned in search, never recommended, and never counted in any pool statistic. This includes a link to a third-party aggregator (Indeed, LinkedIn, ZipRecruiter, Glassdoor, or similar): adding one to your tracker never causes us to fetch it, scrape it, or treat it as a verified posting.

2. How we use your resume data

The pattern: scoring, matching, search, and the anti-fabrication check are our own deterministic code and never leave our systems. Only structuring, rewriting, and review send your content to an AI provider, only when you invoke them, and only the content that feature needs.

3. The anti-fabrication guarantee

This is the core promise of the product, and it is enforced mechanically rather than by policy.

• The rule. A resume produced by Meridino may only contain claims that trace back to what you submitted. It may reorder your words, reword a bullet, and surface a keyword the posting asks for where your own history supports it. It may not add anything else.
• How it is enforced. Every AI-reworded bullet passes through a deterministic gate before it can reach a resume. The gate works by source containment: every content word in the output must trace to your source bullet, to a keyword licensed by your own evidence, or to a small closed list of ordinary function words. On top of that, dedicated checks compare numbers and metrics unit-and-magnitude aware, including spelled-out numbers and multiplier verbs, and detect named entities including short acronyms.
• It fails closed. If the gate detects anything it cannot source, it rejects the entire rewrite and the service keeps your original wording verbatim. A rejection costs you a phrasing improvement. It never produces a fabrication.
• What this means in practice. The service cannot invent an employer, a job title, a degree, a certification, a metric, a percentage, a headcount, a seniority level, a scope claim, or a tool you did not tell us about. If it is not in your history, it cannot appear in your resume.
• Every generated resume carries a change log tracing each claim back to the entry in your own profile that licensed it, so you can audit any bullet yourself.
• What it does not do. The gate governs what a rewrite may contain. It does not verify that what you submitted is true. Accuracy of your own history is your responsibility, and you should read any generated resume before you send it.

4. The AI provider: Anthropic

• Who. Anthropic PBC, United States, the provider of the models behind resume structuring, bullet rewriting, and resume review.
• What is sent. Only what the feature needs. For structuring, the full extracted text of the resume you uploaded, which will include your name, contact details, employers, and dates. For rewriting, your own resume bullets plus the posting's title and the requirement keywords our engine extracted. For review, your resume plus the posting's title and company and our engine's coverage analysis. Content is not redacted before sending, so do not put information in your resume that you would not want a service provider to process.
• When. Only when you use one of those three features. Nothing is sent in the background, on a schedule, or as part of any other feature.
• Free and paid access. An account that has not purchased credits may use AI-based resume structuring exactly once, the one resume upload every account gets, and the ATS coverage check exactly once, which is deterministic and does not use Anthropic. Beyond those two free uses, structuring, rewriting, and review all require purchased credits. We do not claim an unpaid account can never trigger a call to Anthropic; the one free resume upload does.
• On what terms. Anthropic processes the content on our behalf as a service provider under its commercial terms. Those terms provide that Anthropic may not train its models on our customers' content. Anthropic states that it deletes API inputs and outputs within 30 days, except that content flagged as violating its usage policy may be retained for up to two years. We do not control Anthropic's systems; this reflects Anthropic's published terms as of the effective date.
• How to avoid it entirely. Do not use structuring, rewriting, or review. Enter your profile manually and use the deterministic features. Everything else in Meridino works without any AI provider.

5. What we never do

• We never sell your resume, profile, applications, or any other personal data. We have no advertising business and no data-broker relationship.
• We never apply to a job on your behalf. Meridino has no auto-apply feature, no submission integration, and no code path that transmits an application to an employer. You apply yourself.
• We never post, message, comment, or act publicly in your name on any platform.
• We never show your resume, your scores, or your tracker to an employer, a recruiter, or a background-screening company. Meridino is candidate-facing only. Our scores are shown to you, about you, and are never provided to anyone making a decision about you.
• We never use your content to train our own models. We do not train models.
• We never permit an AI provider to train on your content. Anthropic's commercial terms prohibit training on customer content, and we place your content with no provider on any other footing.
• We never run an AI feature on your data unless you invoked it.
• We never send your data, your searches, or anything you type to a search provider.
• We never hold your password, because sign-in is SSO only and no Meridino password exists.
• We never store your card details. Stripe's hosted checkout collects them and we never see them.
• We never store your uploaded resume file. It is parsed in memory and discarded. We do keep the file's name, the text we extracted from it, and the structured reading of that text, in your own account in our database, so that closing the page does not cost you a second AI reading of the same document. There is no store of user files anywhere in the Service.

6. Retention and deletion

The full retention schedule is in Section 9 of the Privacy Policy. In summary:

• Your profile, resume content, generated resumes, scores, applications, and tracker are kept for as long as your account is open and are deleted on a verified deletion request.
• An uploaded resume's file name, extracted text, and structured reading are kept for as long as your account is open, whether or not you go on to confirm it into your record, and are deleted when your account is deleted, along with everything else in the account. We do not delete it on any earlier schedule: we already have the information you gave us, and we can regenerate a resume for you in our own format from what you confirm, so there is no reading-specific expiry to track. The uploaded file itself is never stored at all.
• Search logs including your IP address are kept for security, rate-limiting, and abuse-prevention purposes. We do not currently enforce a fixed deletion schedule for this log.
• Backups roll off within 35 days.
• Billing and payment records are kept for 7 years for tax and accounting purposes.
• Content sent to Anthropic is retained by Anthropic under Anthropic's terms, not ours.

Access, export, correction, and deletion are all handled as a written request, not an in-product self-service control. Write to admin@meridino.com at any time. We will complete a verified request within 45 days, or up to a further 45 days for a complex request, and we will tell you if we need the extension.

7. Your choices

• You decide what goes into your profile. You can edit or delete any entry at any time.
• You decide whether to use an AI feature. They are never automatic.
• You decide whether to apply to anything. We never do it for you.
• There is no subscription to cancel. Meridino is sold as one-time purchases, so nothing recurs and you are never charged again unless you buy again. You can stop using paid features at any time; resumes and drafts already in your account remain until you delete them or delete your account.
• You can request deletion of your account by writing to us, and you can request an export of your data first. See Section 6 for how these requests work.

8. Contact

Questions about how we use data:

Feld & Co LLC
Email: admin@meridino.com
1714 Briar Street A
Austin, TX 78704
United States