SpotFile

Inspiration

We wanted to build a "Google for your local files"

What it does

SpotFile is a desktop application that indexes your local documents and allows you to search through them using natural language.

How we built it

Frontend: React + Tailwind CSS for a snappy, modern user interface.

Core Framework: Tauri, which allowed us to build a tiny, memory-efficient native app using a Rust-based backend instead of a heavy Chromium instance.

The "Brain": We bypassed the traditional Python AI stack in favor of Rust-native tools. We used fastembed-rs for generating vectors and tauri-plugin-llm to run local models (Llama 3.2) for the RAG responses.

Database: A local vector store (Qdrant/VelesDB) stored in the user's hidden AppData folder to ensure persistence and speed.

Challenges we ran into

Deploying our Tauri app with a hosted backend

Accomplishments that we're proud of

Making an app we can use in our daily lives, but most of all the time crunch.

What we learned

Indexing into vectors and RAG

What's next for SpotFile

Adding support for other file types

Built With

• axium
• javascript
• llm
• qdrant
• rag
• rust
• tauri