HR0110 / YOUR DATA

Privacy policy

Effective date: October 10, 2026

This policy covers the HR0110 Chrome extension and its connected service at hr0110.app. HR0110 helps recruiters turn posts from selected Facebook groups into recruitment leads.

1. Who to contact

HR0110 is operated by the product publisher. For privacy questions, access, correction or deletion requests, contact alexdavidov127@gmail.com. You do not need an HR0110 account to email us. Existing users can also use our support chat.

2. Data we process

• Selected Facebook group content: post text, available comments, author names, author profile links, post links and identifiers, and source group links. Text may contain phone numbers, email addresses, cities, work experience, pay expectations and other information the author included.
• Recruitment records: information extracted from collected posts and the records you manage in your workspace.
• Account and service data: your account identifier, name and email supplied through sign-in, workspace membership, session information, and operational records needed to run and secure the service.
• Support messages: messages you choose to send and our replies, associated with your account.
• Local extension data: a next wake-up timestamp supplied by the web application. The extension uses Chrome local storage for this timestamp.

The extension does not read Facebook passwords or cookies, access your private inbox, collect general browsing history, record keystrokes, or track your device's GPS location. It uses the Facebook session already present in your browser. Information such as a city can still be collected when it appears in a post.

3. How collection works

You select Facebook group sources in HR0110 and start collection. The extension opens a browser tab, visits those sources, reads available posts and scrolls to load more content. It returns results to the HR0110 web application, which sends them to the HR0110 server using your signed-in account. Collection does not grant access to groups that your Facebook account cannot access. If Facebook requires sign-in or an identity check, collection stops so you can address it in Facebook.

The extension contains a wake-up mechanism that accepts a timestamp from the application and opens or reloads its tab at that time. It does not independently choose a collection schedule.

4. How we use data

We use data to collect and organize selected posts, avoid duplicates, extract recruitment fields, operate your workspace, authenticate access, respond to support requests, and maintain the service's security and reliability. An extracted lead is a record of a publication, not a verified application or a hiring decision.

5. Sharing and AI processing

Collected records are available to authorized users of the relevant workspace. Support conversations are available to the sender and authorized support administrators.

Hosting and infrastructure providers process data as needed to operate the service. Some features send text to an AI model run by an outside company. We currently use OpenAI for all AI features. Anthropic is our backup provider: if we stop using OpenAI, the same features will run through Anthropic. Vacancy document import works only with OpenAI. Both are US companies, so this processing may take place outside your country, including in the United States. Hosting may also be located outside your country. The extension itself does not download executable code from AI providers.

What the AI model receives, by feature:

• Post analysis in regular collection (automatic): runs in every regular collection, whether you start it, your automation schedule starts it or a connected AI assistant does; there is no separate switch for it. Sent: post text with whatever the author wrote in it (phone numbers, emails, city, experience, pay expectations), the phone numbers and emails already found in that text, and your search scheme descriptions and categories. Not sent: author names, profile and post links, dates, and posts removed earlier by our keyword filter. The model returns its analysis of the post: a short summary, category, city and contacts. It may suggest phone numbers and emails, including ones written out in words; we check them against the post text and mark those not found in it as unverified.
• Post analysis when collecting for vacancies (automatic): runs whenever posts are collected for selected vacancies; there is no separate switch for it, and the keyword filter does not apply in this mode. Sent: the text of every post collected for the task (up to 2,000 posts, up to 24,000 characters each) with whatever the author wrote in it, and the criteria of up to five selected vacancies (the job title and its synonyms, city, schedule, pay, requirements, preferences and quotes from the description). Not sent: author names, profile and post links, dates.
• “Who are we looking for?” (only when you use it): the message you write to describe a vacancy, with anything you include in it.
• Collection plan (only when you use it): the selected vacancies with all their details, including client company, pay, description, requirements, your original request and quotes from the source document; for Facebook groups, only group IDs (no group content), professions and languages. No posts or candidate data. The model only assesses whether each group fits a vacancy by profession and language; our server ranks the groups and makes the selection, and you confirm it.
• Vacancy document import (you start it; OpenAI only): all text extracted from the uploaded file, up to 120,000 characters, including any names or contacts it contains. Vacancies are created only after you confirm them. This import is meant for vacancy documents, not resumes.
• CV Bank (only when you use it): the full text of a resume you upload to a candidate card, up to 70,000 characters: name, contacts, address, work history, education, certificates, pay expectations and anything else in the file. The candidate profile changes only after an HR user reviews the result.
• Search scheme chat and keyword translation (only when you use them): your chat messages and the model's previous replies; for translation, only the keywords and the target language.
• HR assistant (only when you use it): your question, up to 10 previous messages of the conversation, the workspace name, the current screen and the date. The assistant does not read records about candidates, employees or hours worked from your workspace, but anything you type, such as an employee's salary or start date, is sent.

Storage at the provider: we send requests to OpenAI with an instruction not to store them on the OpenAI platform; under its terms, OpenAI may still keep the data for up to 30 days to monitor abuse. How long Anthropic keeps data is set by its own terms. Under their API terms, neither OpenAI nor Anthropic trains its models on this data unless the customer opts in. Provider terms: OpenAI, Anthropic.

Without sending data to AI providers: virus scanning of resumes; extracting text from uploaded files and recognising text in scanned files; the keyword filter in regular collection; match scoring and matching place names to our list of cities; ranking of groups; checking that quotes and data returned by the AI really appear in the source text; collecting posts from Facebook and checking groups in the extension; shifts, hours, payroll calculations, mail and sign-in. These run on our server or in your browser.

People decide: AI output is a suggestion. Vacancy drafts and search schemes are saved only after a person confirms them. A candidate profile changes only after a person confirms the resume data; the resume analysis result is saved right away as an unreviewed version and stays in the workspace even if it is rejected. Post analysis and vacancy matching automatically sort posts into “Qualified”, “Needs review” and “Not suitable”; this sorting helps you work through records, and hiring decisions are made by people. When analysing resumes, vacancy documents and matches, the model is instructed not to guess traits that hiring must not depend on, such as age, sex, religion, origin, health or pregnancy. AI results are stored in your workspace with other records: post analysis and vacancy matches, collection plans, resumes with their extracted text, and HR assistant conversations. You can clear assistant conversations at any time; other results are kept under section 6 or until a deletion request.

AI assistant connector: if you connect HR0110 to ChatGPT, Claude or another AI assistant, that assistant can read and change workspace data within the permissions you approve on the HR0110 consent screen: workspace profile, sources, lead records (post text, AI analysis, author names and links, notes) and, with separate permissions, contact details and your mailbox, including sending emails on your behalf. Post text and AI analysis can contain contact details even without the contacts permission, and the permission to change lead records also lets the assistant change their contact details. The assistant processes what it receives under its own provider's terms, not HR0110's. HR0110 records each time the assistant opens contact details. To end access, remove HR0110 in your assistant's settings and email us: we will revoke access on our side as well.

We do not sell personal data, use it for advertising, or transfer it for unrelated purposes. We may disclose data when required by law or necessary to address security incidents. We do not use or transfer data to determine creditworthiness or for lending.

6. Storage and deletion

The extension does not keep a persistent archive of collected posts. The connected service stores workspace records on its server. Retention depends on the workspace plan and whether a post is linked to a lead: eligible older, unlinked post content is redacted by the retention process, while linked posts and recruitment records remain available for workspace use. Identifiers, links and duplicate-detection metadata can remain after content redaction. Vacancy matches keep their own copy of the post text, author name and post link; the retention process does not redact this copy, and it remains until deleted on request. This process is not complete account deletion.

To request access, correction or deletion of an account, workspace records, support messages, or a post concerning you, email alexdavidov127@gmail.com. Include enough information to identify the record, such as its post link; do not send passwords or identity documents unless a verification method has been agreed. We may need to verify your identity or authority before acting.

Deleted data may remain temporarily in restricted backup copies until they expire or are rotated. Records may be retained where necessary to meet legal obligations or resolve disputes. Uninstalling the extension stops its browser activity but does not delete records already stored in HR0110.

7. Permissions and security

• scripting: runs the packaged extraction script in collection tabs and the packaged communication bridge in HR0110 tabs.
• tabs: opens, navigates and closes collection tabs and locates or reloads the HR0110 tab.
• alarms: wakes the application at a timestamp it supplies.
• storage: stores the next wake-up timestamp locally.
• https://*.facebook.com/*: accesses selected group pages through your existing browser session.
• https://hr0110.app/*: exchanges predefined commands, progress and collection results with HR0110.

All executable extension code is included in the extension package. The service uses HTTPS and authenticated access to workspace and support data. Please do not include passwords, API keys or unnecessary sensitive information in support messages or collected records.

8. Chrome Web Store Limited Use

HR0110's use and transfer of data obtained through the extension complies with the Chrome Web Store User Data Policy, including its Limited Use requirements. Data is used and shared only to provide or improve the extension's stated recruitment collection functionality, or for permitted security and legal purposes.

9. Changes

We will update this page when our practices change and revise the effective date. Material changes requiring additional notice or consent will be communicated before the changed processing begins.