Inspiration

After months of applying to jobs with degrees, projects, and certifications, all I got back was silence. Eventually I burned out and stopped applying altogether. The problem wasn’t skills. It was the process: repetitive, unclear, and mentally draining.

At the beginning of this hackathon, Applybot Pro was simple. It only analyzed CVs against job descriptions and generated better applications. But while testing it, I noticed something deeper: people don’t actually know when they should apply. Most applications fail before they’re even sent.

So the project evolved. Instead of just helping people apply better, it started helping them decide whether to apply at all.

That’s when the briefing page and browser extension were added, turning it from a resume helper into a decision-making assistant. It evolved to a career pilot.


What it does

Applybot Pro is an AI-powered job application assistant designed to improve success rate, not application volume. It searches jobs across multiple platforms, evaluates how well you match a role, and only proceeds if the fit makes sense. Before applying, a briefing page shows:

  • real match score
  • skill gaps
  • recommended improvements
  • whether the application is worth your time

If you continue, it:

  • generates tailored cover letters and outreach emails
  • tracks applications
  • prepares interview questions based on your CV and that specific job
  • auto-fills forms using a browser extension (optional you stay in control)

The goal isn’t automation. The goal is intelligent effort.


How we built it

Frontend: React + TypeScript (Vite) Backend: Python APIs + Supabase database AI: Gemini reasoning for job matching, document tailoring, and interview prep Extension: browser assistant for guided applications and autofill


Challenges we ran into

  • Normalizing job data from different platforms
  • Scoring candidate–job fit honestly without misleading users
  • Avoiding spam-like automated applications
  • Designing feedback that helps rather than discourages
  • Balancing automation with user control and trust

Accomplishments that we're proud of

  • Built a full decision-to-application workflow, not just a generator
  • Honest mismatch detection (“don’t apply” guidance)
  • Context-aware cover letters and emails
  • Interview preparation grounded in the actual application sent
  • Extension-assisted applications instead of blind autofill

What we learned

  • Job seekers don’t need more applications; they need better targeting
  • Transparency builds trust more than convenience
  • People value guidance more than automation
  • The hiring process filters effort, not just talent

What's next for Applybot Pro

  • More job platform integrations
  • Explainable match scoring
  • Resume improvement feedback loops
  • Analytics on what actually leads to interviews
  • Recruiter-side insights (optional)
  • Eventually: a continuous AI career companion, not just an application tool

Design assistance: UI styling iterations were supported using Lovable.dev during development.

Built With

Share this project:

Updates