AI Resume Toaster

Inspiration

Resume feedback is usually vague or overly polite. Friends say “looks good,” recruiters say nothing, and applicants are left guessing why their resume didn’t work. We wanted to build something that gives the kind of feedback people rarely get: honest, specific, and immediately useful. The idea was to combine tough love with real improvement — roast the resume, then actually fix it.

What it does

AI Resume Toaster analyzes resume bullet points, critiques them with clear and blunt feedback, and then rewrites them into stronger, more impactful versions. Users can control the roast intensity, making the feedback anything from recruiter-safe to brutally honest. The goal is to turn vague bullets into concrete, results-driven statements.

How we built it

The project is built with SvelteKit for a fast and interactive frontend and deployed on Vercel with a free public domain. AI calls are handled server-side to keep API keys secure. We use carefully designed prompts and few-shot examples to teach the model how to judge resume quality, not just rewrite text.

Challenges we ran into

One of the biggest challenges was getting the AI to be honest without becoming uselessly harsh or generic. Early versions were either too polite or too repetitive. We also ran into deployment issues around environment variables, which required moving secrets out of local files and into production settings. Balancing humor, clarity, and real value took several iterations.

Accomplishments that we're proud of

Built a fully deployed, shareable app with a clean UI Created feedback that feels opinionated and human, not generic Successfully separated critique and rewrite logic for clearer results Shipped a complete end-to-end experience under time pressure

What we learned

We learned how important prompt design is for controlling tone and consistency in AI outputs. We also gained hands-on experience deploying a real application with proper environment handling and server/client separation. Most importantly, we learned that users value clarity and honesty more than vague positivity.

What's next for AI Resume Toaster

Next steps include resume scoring, side-by-side before/after comparisons, more customization over tone, and exportable rewritten resumes. We also want to expand beyond resumes into cover letters and LinkedIn summaries.

Built With

• html
• javascript
• sveltekit
• typescript