Youxai Job Copilot

YouxAI Job Copilot is a browser side panel designed to simplify the job application process through a hybrid AI architecture. By combining local processing with cloud sync, it offers a balance of privacy, speed, and data organization.

Technical Architecture: Local AI
Unlike standard extensions that rely entirely on cloud APIs, YouxAI uses a Local LLM to process sensitive job data directly in your browser.

Privacy: Initial job analysis and match scoring happen on your machine.
Efficiency: Processing locally reduces latency and allows for frequent analysis without external API costs.
PII Management: If you choose to sync your applications to your YouxAI dashboard, we prioritize stripping identifiable information (PII) during the transfer.
Automation Modes
Our automation system is built to handle different application workflows:

Autonomous Agent: Specifically tuned for LinkedIn Easy Apply to navigate multi-step modals and submit automatically.
Draft & Flow: For complex enterprise forms (Workday, Greenhouse, Lever), the AI maps your profile data to fields in the background, allowing you to review the draft before submission.
Core Features
Job Capture: Detect and save structured data from LinkedIn, Indeed, Lever, Greenhouse, and Workday.
Match Scoring: Immediate feedback on how well your profile aligns with a job description.
Magic Autofill: Uses your Master Profile data to generate relevant drafts for application questions.
Unified Dashboard: Sync your activities to the YouxAI web app for long-term tracking.

v4.1.3 Release Notes:
● Minor bug fixes in linkedin context detection
● Emergency auth fix

v4.1.2 Release Notes:

● Improved apply-flow reliability, especially for Jobvite multi-step journeys and consent/application transitions.
● Strengthened autofill adapter resolution using portal + hostname + URL hints for more deterministic behavior.
● Added smarter Jobvite handling: resume upload candidate ranking, radio-group question inference, and safer fallback paths.
● Improved form detection and submission reliability with retry scans and safer submit-listener handling.
● Added match-score stability improvements via better job identity matching and cache reuse.
● Improved Discover quality by deduplicating jobs and prioritizing higher-quality ATS listings.
● Upgraded Dashboard and Applications UX: better empty/loading states, URL-synced filters/sorting, accessibility polish, and extension-status hook reuse.
● Updated aggregation quality/dedup logic in Supabase functions to reduce duplicate jobs and improve source scoring.