This is a submission for the Hacktoberfest Open-Source AI Challenge Week 3: Slow and Steady.
What I Built
I built Sloth Sorter, a local AI-powered tool that helps organize files with less manual effort.
Instead of opening every file and deciding where it belongs, users can select a folder, scan its contents, and let AI suggest appropriate categories and more meaningful filenames.
The tool can:
Scan files in a selected directory.
Suggest categories and filenames using local AI.
Preview proposed changes before applying them.
Move files into category folders.
Log file operations to support undo functionality.
The goal is simple: spend less time organizing files and more time on productive tasks!
Demo
GitHub Repository: https://github.com/Abhinav55849/sloth-sorter
The repository contains the source code and example files. The application currently runs locally using Streamlit and Ollama; it is not a publicly deployed web application.
Code
GitHub: https://github.com/Abhinav55849/sloth-sorter
The repository includes:
The Python application.
AI-powered file classification logic.
File scanning and moving modules.
Example files for testing.
Automated tests.
How I Built It
I built Sloth Sorter using Python and Streamlit, with Ollama to run Gemma 3 locally.
Tech Stack
Python: Application logic and file handling.
Streamlit: Interactive user interface.
Ollama: Local model inference.
Gemma 3: AI-powered file classification and filename suggestions.
Pytest: Automated testing.
The application scans a selected folder and prepares file information for classification. The local AI model then suggests categories and filenames, which users can review before applying changes.
I also implemented fallback classification logic for situations where the AI response is unavailable or invalid.
Why Does Open Innovation Matter?
Open innovation makes it easier to experiment with AI, build useful tools, and learn without depending entirely on hosted APIs.
Using an open-weight model locally offers several potential benefits:
More control over data: Local inference can reduce the need to send file information to external AI services.
Accessibility: Developers can experiment without necessarily paying for a hosted model API.
Flexibility: Local setups make it possible to experiment with different compatible models.
Learning and experimentation: Developers can explore how AI can be integrated into practical applications.
This project helped me understand how local AI can work alongside traditional Python file operations to solve an everyday problem.
My Agent Session
I don't currently have a DevRelay session link to share, so this optional section has been omitted.
The Lazy Payoff
Sloth Sorter aims to reduce the repetitive effort involved in manually organizing files.
Instead of reviewing every file and deciding where it belongs, users can review AI-generated suggestions and apply the changes they want.
I built this project to explore how local AI could simplify this workflow. I haven't measured the actual time saved yet, so I won't claim a specific figure.
Prize Categories
I'm submitting Sloth Sorter to the Hacktoberfest Open-Source AI Challenge — Week 3: Slow and Steady.
Conclusion
Thanks for checking out Sloth Sorter! 🦥
I'd appreciate feedback, suggestions, and contributions from the open-source community.
GitHub Repository: https://github.com/Abhinav55849/sloth-sorter
Previous project write-up: https://dev.to/abhinav55849/sloth-sorter-a-local-ai-that-organizes-files-so-you-dont-have-to-2lii
Top comments (0)