DevpostResume.AI

DevpostResume.AI is a platform to convert your hackathon demo video into past-able devpost write-up, it also allows users to customize resume based on his selection of uploaded demo videos.

Comment 1

Inspiration

Writing a strong Devpost submission or resume after a hackathon is often harder than building the project itself. Many teams finish demos exhausted, then struggle to turn a long demo video into clear, well-structured written content. We were inspired to reduce this friction by transforming spoken project demos into high-quality Devpost writeups and resumes automatically, so builders can focus on building rather than rewriting.

What it does

DevpostResume.AI converts a project demo video into structured, paste-ready Devpost sections and a professional one-page resume. Users upload a demo video, and the platform transcribes it, extracts key ideas, and generates organized sections such as Inspiration, What it does, and Challenges. Users can also select past projects and generate a resume in Markdown or PDF format, making it easy to reuse hackathon work for job applications.

How we built it

We built a full-stack web application using a Next.js frontend and a FastAPI backend. Video uploads are transcribed server-side, then processed with Gemini to generate structured Devpost sections and resume content. Generated projects and resumes are stored in MongoDB, while finalized PDFs are uploaded to Cloudflare R2 for scalable storage. Phantom Wallet authentication is used to associate projects and resumes with a user’s wallet address.

How we built it (Tech Stack)

DevpostResume.AI is built as a full-stack application that combines AI services, blockchain-based identity and payments, and cloud storage.

We use Phantom Wallet for authentication and user identity, allowing each project and resume to be associated with a unique wallet address without requiring traditional accounts or passwords. Payments are handled on Solana Devnet, where users pay per resume generation using a lightweight pay-as-you-go transaction model.

The frontend is built with Next.js, providing a responsive interface for uploading demo videos, selecting past projects, and generating resumes. The backend is powered by FastAPI, which handles video uploads, orchestration of AI services, payment verification, and data storage.

For AI processing, we use Whisper to transcribe demo videos into text. The transcripts are then processed with Gemini, which converts raw speech into structured Devpost sections and resume-ready summaries while ensuring that no information is invented beyond what is stated in the demo.

Generated projects and metadata are stored in MongoDB, allowing users to revisit and reuse previous submissions. Finalized resume PDFs are uploaded to Cloudflare R2, providing scalable and cost-efficient object storage with secure access via signed URLs.

Together, this stack enables a seamless pipeline from demo video → structured writeup → professional resume → downloadable PDF.

Tech Stack Summary

Frontend: Next.js

Backend: FastAPI (Python)

Authentication: Phantom Wallet

Payments: Solana Devnet (pay-as-you-go model)

Transcription: Whisper

AI Writing & Summarization: Gemini

Database: MongoDB

File Storage: Cloudflare R2

Tech Stack Diagram

Tech Stack

[ User Browser ]

  |

  v

[ Next.js Frontend ]

  • Video Upload

  • Project Selector

  • Resume Preview

    |

    | Phantom Wallet (Auth + Payment)

    v

[ Solana Devnet ]

  • Pay-as-you-go (0.01 SOL)

    |

    v

[ FastAPI Backend ]

  • Verify payment

  • Handle uploads

  • Orchestrate AI

    |

    +--> [ Whisper ]

    | - Video → Transcript

    |

    +--> [ Gemini ]

    | - Transcript → Devpost sections

    | - Projects → Resume

    |

    +--> [ MongoDB ]

    | - Projects

    | - Wallet ownership

    |

    +--> [ PDF Generator ]

    | - Resume → PDF

    |

    +--> [ Cloudflare R2 ]

         - Resume PDF Storage

Challenges we ran into

One challenge was ensuring that the generated content strictly reflected only what was stated in the demo video, without inventing features or results. Another challenge was formatting resumes and PDFs in a way that looked professional while remaining fully automated. We also had to carefully manage file uploads, PDF generation, and cloud storage integration across multiple services.

Accomplishments we're proud of

We’re proud that DevpostResume.AI produces clean, structured Devpost writeups that are immediately usable without editing. The resume generation feature allows users to select multiple projects and create a polished one-page PDF in seconds. We also successfully integrated wallet-based identity, cloud storage, and AI content generation into a cohesive workflow.

What we learned

Through this project, we learned how much value clear documentation and presentation add to technical work. We gained experience building reliable AI-assisted writing tools, handling media pipelines, and designing user flows that reduce friction at the end of a hackathon. We also learned how to integrate multiple cloud services into a single production-ready system.

What's next

Next, we plan to add multiple resume styles, editable headers, and project-specific tailoring for different roles. We also want to introduce team collaboration features and expand support for additional formats such as cover letters and portfolios. Our goal is to make DevpostResume.AI a standard post-hackathon workflow tool for builders everywhere.

Built With

Share this project:

Updates