JAI

An AI-powered job application agent that tailors resumes, autofills applications, tracks outcomes from emails, and helps you reach the right hiring managers.

Comment

🚀 Inspiration

Applying for jobs today is repetitive, time-consuming, and mentally exhausting. Candidates spend hours rewriting resumes, copying the same information into different application forms, and manually tracking submissions, only to repeat the process for the next role.

We wanted to build something we personally needed: an intelligent system that bridges the gap between a single Master Resume and hundreds of job-specific applications. JAI was inspired by the idea of turning the job application process from a manual chore into an automated, intelligent workflow, while still keeping the user in control.

🧠 What it does

JAI (Job Application Intelligence) is an AI-powered ecosystem that automates and optimizes the end-to-end job application workflow.

It consists of two tightly integrated components:

Chrome Extension (The “Actor”)

  • Scrapes job descriptions directly from job postings (LinkedIn, Indeed, company career pages).
  • Uses Google Gemini to tailor resume content specifically for each job.
  • Automates Overleaf to generate clean, professional, ATS-friendly PDF resumes from LaTeX.
  • Autofills job application forms (name, email, links, experience fields) on company websites.
  • Provides visual feedback showing which fields were successfully filled and which require review.

Web Dashboard (The “Brain”)

  • Stores a user’s Master Resume and profile information.
  • Manages personal details used for autofill.
  • Acts as the central hub for AI processing and resume generation.
  • Handles authentication and session management shared with the extension.

Together, JAI turns a job posting into a tailored resume and a partially completed application in just a few clicks.

🛠️ How we built it

We designed JAI with a clear separation of responsibilities:

  • Chrome Extension (Manifest V3)

    • Content scripts extract job descriptions and interact with application forms.
    • Background service worker orchestrates API calls and communicates with the backend.
    • Popup UI provides controls and status feedback.

  • Web Dashboard

    • Built using Next.js, MongoDB, and Tailwind CSS.
    • Stores the Master Resume and user profile.
    • Uses Google Gemini API to intelligently rewrite resume content based on job descriptions.
    • Generates structured resume output that is converted into LaTeX.

  • AI Integration

    • Gemini is used for reasoning, not just text generation.
    • Resume tailoring follows strict formatting constraints to preserve layout and ATS compatibility.
    • Autofill logic prioritizes deterministic rules first, with AI used only as a fallback for ambiguous fields.

  • Automation

    • Overleaf automation injects generated LaTeX, compiles PDFs, and prepares resumes for download.
    • Autofill logic safely fills common application fields while skipping sensitive or subjective questions.

⚠️ Challenges we ran into

  • Browser security constraints limited what could be automated safely.
  • Maintaining one-page resume layout while modifying content with AI was challenging.
  • Designing autofill logic that avoided sensitive or demographic fields required careful filtering.
  • Synchronizing state between the extension and the web dashboard added architectural complexity.

🏆 Accomplishments that we're proud of

  • Built a complete end-to-end system rather than a standalone demo.
  • Successfully automated resume tailoring without breaking LaTeX formatting.
  • Integrated browser automation, AI reasoning, and a cloud backend into one workflow.
  • Designed autofill that is transparent, safe, and user-controlled.
  • Created a system that feels production-oriented rather than a hacky prototype.

📚 What we learned

  • AI is most effective when combined with deterministic logic.
  • Browser extensions require careful architectural boundaries.
  • Separation of “thinking” (AI/backend) and “acting” (extension) improves reliability.
  • User trust is critical when building automation tools.

🔮 What's next for JAI

  • Expand autofill support across major ATS platforms (Greenhouse, Lever, Workday).
  • Add email intelligence to track interviews, rejections, and offers automatically.
  • Introduce personalized recruiter and hiring-manager outreach assistance.
  • Improve resume version tracking and application analytics.
  • Package JAI as a deployable tool for students and early-career professionals.

Built With

Share this project:

Updates