Seamstress

Inspiration

We all know the job market is abysmal. At the same time, students and professionals are busy balancing classes, internships, and life, and don’t always have the time to constantly rewrite and tailor their resumes. As two students, some time you just don't have the energy.

We were inspired by the personal frustration of spending hours manually adjusting resumes for every job application and constantly making sure we didn't mess up wording or spacing while we are at it.

Our goal was to create something that reduces friction in the job search process and empowers users to apply with confidence.

How we built it

We built our application using a React frontend and a FastAPI backend deployed on DigitalOcean App Platform. All generated LaTeX files, static assets, and uploaded resumes are stored in DigitalOcean Spaces (S3 compatible object storage), enabling scalable file management and secure presigned URL access.

When a user uploads a resume, the backend extracts the raw text and converts it into a structured JSON schema validated with Pydantic. We then integrate the Google Gemini API in two primary workflows:

• Resume parsing and AI powered editing, where Gemini rewrites content for clarity and professionalism without inventing new facts. • Job matching and tailoring, where Gemini analyzes job descriptions and salary preferences to generate targeted resume improvements.

All AI responses are strictly validated against our schema before being rendered into LaTeX for professional PDF generation. The architecture is modular and API driven, cleanly separating storage, AI processing, validation, and rendering to ensure reliability, scalability, and maintainability.

What we learned

Both members of our group had no experience using the Gemini API or DigitalOcean, so in every sense we learned as we went! One of the challenges we faced was out Gemini Key running out of use too often so we have to manually cycle in new keys.

What's next for Seamstress

ext, we plan to expand beyond resume editing into smarter job decision tools. Our original roadmap included a location-based cost-of-living feature, and we’d like to integrate external APIs to compare salaries against real expenses by city. Users could factor in commute distance, car ownership, and estimated gas costs to understand true take home value , not just salary numbers.

We also want to further refine the AI’s tone control, making it sharper and more dynamic while maintaining strict factual accuracy. Finally, we plan to broaden the platform beyond tech focused resumes to support multiple industries and formats.

Built With

• digitalocean
• digitaloceanbuckets
• gemini
• python
• rapidapi
• theirstack