Jobs matched to what's proven

How matching works

Every job description and every profile is converted into a semantic embedding — a representation of meaning, not keywords. Matching compares the two, so "built REST APIs in FastAPI" matches a posting that says "backend services in Python" even though the words differ. Each match gets a role-fit score, and your best matches surface on your dashboard ("3 perfect-fit roles").

The jobs feed

• Roles are aggregated from multiple job boards and refreshed continuously.
• Each listing carries an AI summary so you can scan fit in seconds.
• Filters cover role type, location/remote, and experience level.
• Salary is shown when the source discloses it — we don't invent numbers.

Applying

Applying through the platform earns 75 XP and tracks the application status in one place. Getting shortlisted for an interview earns 300 XP — outcomes count more than volume. When you apply, the employer sees your public profile card alongside your application.

Improving your matches

• Verify skills — matching weights verified skills above self-reported ones.
• Keep your experience section specific; embeddings match on meaning and detail.
• Ask Vanilla to review a match before applying — it can see both the role and your profile.