Inspiration
As students, we found ourselves spending hours manually tailoring cover letters for every application. We realized that while AI can solve this, most tools require uploading sensitive resumes to the cloud/apps outside your Hardware. We wanted a solution that prioritized privacy-first automation.
What it does
Localcraft.io is a "Privacy-by-Design" career suite. A browser plugin scrapes job descriptions (JDs) from boards, which a local desktop app then processes using Local LLMs to generate high-quality, customized cover letters—ensuring your PII never leaves your machine.
How we built it
We built a cross-platform pipeline integrating a browser extension for data extraction and a local Python/Streamlit app for the inference engine. The system utilizes local LLM integration to handle the synthesis of customised cover letters with real-time JD data. Our Data API's are hoster on render.com
Challenges we ran into
Ever job board has different structure , so scraping them need to address that structure . Size of the model and Output determines the time to delivery .
Accomplishments that we're proud of
We successfully bridged the gap between modern AI and classic Software Engineering. We proved that high-utility AI doesn't require a cloud-first, data-harvesting model, achieving absolute data sovereignty for the user. Simple Everyday jobs require carefully crafted simple everyday solutions
What we learned
Simple tools go a long way in solving everyday problems. We learned that while data is the fuel for modern AI, solid SE tools (like plugins and local APIs) still form the essential backbone that makes AI truly accessible and secure.Also leverage the modern Platforms for most Backend tasks like deployment
What's next for LocalCraft.io-UMD
While our API is Deployed , it's on a Free tier , Plan to move to a real production tier . In terms of App feature we want it to expand beyond Cover letters , into things like Skill gap analysis , match % etc .
Log in or sign up for Devpost to join the conversation.