This is a submission for the "Hacktoberfest Open-Source AI Challenge Week 3: Slow and Steady" (https://dev.to/challenges/hacktoberfest-week3-2026-10-19)
What I Built
Meet LazySort, a local AI-powered file organizer that takes care of the digital clutter we keep postponing.
Downloads folders are full of random PDFs, screenshots, documents, images, code files, and files with names that make absolutely no sense. Organizing everything manually is boring, so why not let AI handle it?
LazySort is designed to scan a selected folder, understand what files contain, suggest meaningful filenames, and organize them into categories such as Documents, Work, Images, Code, and Personal.
The goal is simple: spend less time managing files and more time doing things that actually matter.
Key features:
- Automatically categorize files using AI.
- Suggest meaningful filenames instead of confusing defaults.
- Organize cluttered folders into a clean structure.
- Preview proposed changes before applying them.
- Keep files and sensitive information on the local machine.
The intended demo workflow is straightforward: select a messy folder, let the local AI analyze the files, review the suggested organization, and apply the changes.
Code
The project will be open source, allowing developers to inspect the implementation, customize the categorization rules, and contribute improvements.
How I Built It
LazySort is designed around open-source AI rather than depending entirely on a proprietary cloud API.
Planned technology stack:
- Python: Core application logic and file operations.
- Ollama: Local model execution and inference.
- Qwen or another suitable open-weight model: Understanding file content and generating useful filenames and categories.
- Local file system: Reading selected files and organizing them into folders.
- Streamlit: A simple interface for reviewing suggested changes.
The workflow is:
- The user selects a folder.
- The application identifies supported file types.
- Relevant text or metadata is extracted locally.
- The open-weight model suggests a category and an appropriate filename.
- The application presents a preview of the proposed changes.
- The user approves the changes before files are moved or renamed.
The application is designed to keep the model's analysis local, with safeguards such as previewing changes and avoiding overwrites.
Why Does Open Innovation Matter?
For a tool that handles personal documents, privacy matters just as much as convenience.
A closed AI API could provide file classification, but it may require sending document contents to an external service and could introduce ongoing usage costs.
An open-weight model running locally offers a different approach.
- Privacy by design
Personal documents can contain financial information, work files, private notes, and other sensitive content. Local inference can keep document contents on the user's machine instead of sending them to a third-party AI API.
- Freedom to choose the model
The project can use different compatible open-weight models depending on the available hardware, speed, and classification quality. Developers can experiment without being locked into one provider.
- Reduced recurring costs
Once the software and model are downloaded, local inference can run without per-request API charges. Hardware, electricity, storage, and setup still have costs.
- Offline capability
After setup and downloading the required model, the core workflow can be designed to run without an internet connection.
- Customization and transparency
Developers can modify prompts, categorization rules, and application logic to suit their own workflow. They can also inspect the code and adapt the tool rather than relying entirely on a closed service.
For this use case, open innovation isn't just about using a different model. It makes a more private, customizable, and potentially cost-effective approach to everyday automation possible.
The Lazy Payoff
Imagine downloading 50 files after a week of work and having to sort each one manually.
LazySort aims to reduce that repetitive work to a quick review and approval process. Instead of opening every file, deciding where it belongs, and renaming it yourself, you let a local AI assistant prepare the organization for you.
The intended payoff is less repetitive work, a cleaner workspace, and more time for the things you would rather be doing.
Actual time savings and classification accuracy will need to be measured during testing.
Built for the lazy, powered by open innovation.
Why spend your evening organizing files when you could be relaxing like a sloth? 🦥
Top comments (0)