GIF
demo video

Inspiration

Job hunting is broken. The average applicant spends weeks copy-pasting the same information into slightly different forms, rewriting the same resume for every posting, and firing applications into the void. We watched friends apply to 200+ roles by hand — and burn out before they ever got an interview. So we asked a simple question: what if applying to a job took one click instead of one hour? That's OneClick Apply.

What it does

OneClick Apply turns a single resume into a stream of tailored, ready-to-submit job applications:

Finds matching roles based on your skills, experience, and preferences
Tailors every application — rewriting your resume and cover letter to match each posting's keywords and requirements
Submits for you, so you spend your time interviewing instead of filling out forms

Upload once, and let the agent do the repetitive work.

How we built it

Frontend: Next.js + Tailwind CSS for a fast, responsive UI
AI layer: an LLM-powered agent that parses the resume, reads each job posting, and generates a customized application per role
Infrastructure: provisioned through Stripe Projects, wired up directly from our editor during the hackathon
Deployment: Vercel

The flow is: parse resume → match jobs → tailor content per posting → submit.

Challenges we ran into

Tailoring without hallucinating — the model had to rephrase and emphasize real experience for each posting without inventing skills the applicant doesn't have.
Structured output from messy inputs — resumes and job listings come in wildly different formats, so reliable parsing took real effort.
New tooling under time pressure — we set up our stack with Stripe Projects live during the event, learning the workflow as we built.

Accomplishments that we're proud of

Shipped a working end-to-end flow — from resume upload to a tailored, submit-ready application — within the hackathon window.
Built tailoring that stays grounded in the applicant's real experience instead of inventing qualifications.
Provisioned and deployed a full stack as a small team using tooling we'd never touched before.

What we learned

Prompt design matters more than model size for keeping tailored output honest and on-target.
A tight "one click" UX is harder than the AI behind it — cutting friction is the real product.
Modern agent tooling lets a small team provision and ship a full stack in a single hackathon.