Data Usage Policy

1. Overview

JobCompanion is a data-driven platform. To deliver accurate job matching and a personalized application experience, we process several categories of user data using a combination of AI models, structured databases, and algorithmic pipelines. This policy explains exactly what data is processed, how it flows through our systems, and the controls you have over it.

2. Resume Ingestion and AI Parsing

2.1 Upload and Storage

When you upload a resume, the file is transmitted over an encrypted HTTPS connection and stored in Supabase object storage. The file is associated with your authenticated user account and is not accessible to other users or the public.

2.2 AI-Powered Text Extraction

Resume content is extracted and parsed using AI language models (including OpenAI GPT-series APIs). The extraction pipeline performs the following operations:

• Raw text extraction from PDF, DOCX, or plain text files
• Structured parsing of work experience (company, role, dates, responsibilities)
• Education history extraction (institution, degree, graduation year)
• Skills identification and normalization (technical skills, soft skills, tools, frameworks)
• Contact information extraction (name, email, phone — stored only for profile use)
• Seniority level inference based on years of experience and role titles

2.3 Structured Profile Creation

The extracted data is stored as a structured profile in our database. This structured profile — not the raw resume file — is the primary data source used for job matching. You can review and update your structured profile at any time through the Profile & Preferences section.

3. Job Matching Algorithm

3.1 Job Data Sources

JobCompanion aggregates job listings from multiple public job data providers (including but not limited to RemoteOK and Arbeitnow). Job data is ingested, normalized, and stored in our database. We do not scrape employer websites directly.

3.2 Matching Process (High-Level)

The matching pipeline operates as follows:

• Skill Overlap Scoring: Your extracted skills are compared against required and preferred skills listed in each job description. A weighted overlap score is computed.
• Role Title Alignment: Your current and target job titles are compared against job titles in listings using semantic similarity.
• Preference Filtering: Jobs are filtered based on your stated preferences (location, remote/hybrid/on-site, seniority level, industry).
• Recency Weighting: More recently posted jobs receive a higher relevance weight to surface timely opportunities.
• Match Score: A composite match score (0–100) is computed for each job and used to rank listings in your dashboard.

The matching algorithm is continuously refined based on aggregated, anonymized user interaction data (e.g., which jobs users apply to vs. dismiss). Individual user data is never used to train AI models without explicit consent.

3.3 No Automated External Submission

4. Application Tracking Data

JobCompanion maintains a structured log of your application activity, including:

• Jobs added to your application queue
• Application status updates (applied, interviewing, offer, rejected, withdrawn)
• Timestamps of status changes
• User notes and annotations on specific applications
• Application mode used (manual, human-assisted, or AI-assisted)

This data is used exclusively to power your application history dashboard and is not shared with employers or third parties.

5. System Logs and Diagnostics

We maintain system-level logs for the following purposes:

• Error Logging: Application errors and exceptions are logged to diagnose and fix platform issues. Logs may contain request metadata but not sensitive personal data.
• Performance Monitoring: API response times and system health metrics are tracked to maintain platform reliability.
• Security Logging: Authentication events (login, logout, failed attempts) are logged for security monitoring and fraud detection.

System logs are retained for a maximum of 90 days and are then automatically purged. Logs are not used for user profiling or marketing.

6. AI Model Data Usage

When your resume or job data is sent to OpenAI APIs for processing:

• Data is transmitted over encrypted connections to OpenAI's API endpoints
• OpenAI's API usage policy (as of the date of this policy) does not use API-submitted data to train their models by default
• We do not send your name, email, or contact information to AI APIs — only resume content and job description text
• AI processing is stateless — no conversation history or user context is retained between API calls

7. User Control Over Data

You have the following controls over your data within the Platform:

• Resume Management: You can upload, replace, or delete your resume at any time. Deleting your resume removes the file and triggers re-processing of your profile.
• Profile Editing: Your structured profile (skills, preferences, job history) can be edited directly in the Profile & Preferences section.
• Application History: You can view, annotate, and manage your application history. Individual application records can be removed.
• Account Deletion: Deleting your account permanently removes all associated data, including your resume, profile, and application history, within 30 days.

8. Data Minimization Principle

We adhere to a data minimization philosophy: we collect and retain only the data necessary to provide the Platform's core functionality. We do not collect demographic data (race, gender, age, religion) and do not use such data in our matching algorithms. We do not build behavioral advertising profiles or sell data to data brokers.