What I Built
Meet SlothDesk AI 🦥 — a local-first file organization assistant that helps you spend less time sorting files and more time doing what actually matters.
Downloads folders have a habit of becoming digital junk drawers. Documents, images, archives, and miscellaneous files pile up until finding anything becomes a task of its own.
SlothDesk AI helps bring order to that chaos.
It scans a folder, analyzes file metadata, and builds an organization plan that you can review before applying any changes. It can also use optional, locally running Ollama models to suggest file categories and more meaningful filenames.
The principle behind it is simple: AI should do the tedious thinking, but you should stay in control.
What makes it different?
🌿 Local-first: Work with files on your own device.
🧠 Optional local AI: Get organization suggestions through Ollama without depending on a paid AI API.
🛡️ Human approval: Review the proposed changes before applying them.
📁 Safer organization: Designed to avoid accidental overwrites and file deletion.
⚡ Simple workflow: Scan, review, and organize through a Streamlit interface.
It's for anyone whose Downloads folder needs a little less chaos — students, developers, researchers, and anyone who would rather not sort files manually.
Demo
🎥 Demo video: Watch my SlothDesk AI walkthrough
https://drive.google.com/file/d/1WQiO4eIE6i3EJqLnfjQQDny6ANSVsF87/view?usp=drivesdk
The walkthrough demonstrates the app's interface and file-organization workflow.
Code
💻 GitHub repository:
https://github.com/SM01010011/SlothDesk-AI
The project is built as an open-source Python application. Explore the repository for the implementation, setup instructions, and tests.
How I Built It
SlothDesk AI combines a lightweight Python application with an optional local AI suggestion engine.
Tech stack:
Python — core application and file-organization logic
Streamlit — interactive user interface
Pandas — tabular presentation of scanned file information
Ollama — optional local inference for AI-powered organization suggestions
Pytest — automated testing
The workflow separates planning from execution. The application scans the selected folder and builds an organization plan first. The user can review that plan before choosing to apply it.
The AI component is optional. When enabled, it helps suggest categories and filenames using file metadata rather than sending file contents to the suggestion engine. This keeps the project focused on practical assistance without giving an AI model unrestricted control over the filesystem.
I wanted the experience to feel useful even without a cloud AI subscription. Local inference makes that possible for users with a compatible Ollama setup and model.
Why Does Open Innovation Matter?
For a project like SlothDesk AI, open innovation isn't just about making the source code visible. It changes who can use, inspect, adapt, and improve the tool.
Using open-source tools and local inference makes experimentation possible without requiring a paid proprietary AI API. Developers can inspect the implementation, run it in their own environment, test different compatible models, and adapt the workflow to their needs.
It also makes privacy-conscious experimentation more accessible. Instead of building the entire experience around sending data to a remote service, SlothDesk AI can use local models for its optional suggestions.
There is still work to do: local inference depends on the user's hardware and installed model, and AI-generated suggestions should always be reviewed. But open tools make it possible to build and improve this kind of practical assistant with fewer external dependencies.
That's the kind of open innovation I wanted to explore: use AI to reduce repetitive work without sacrificing user control.
The Lazy Payoff
The goal is to reduce the repetitive effort involved in sorting a cluttered folder.
Rather than manually deciding where each file belongs and thinking up better names one at a time, users can scan a folder, review suggested organization, and approve a plan.
I haven't measured a verified time-saving figure yet, so I won't claim a specific number of minutes saved. The next step is to test it on real, cluttered folders and measure how much manual effort the workflow actually removes.
Slow and steady — build something useful, test it honestly, and keep improving it. 🌱
Prize Categories
This submission is entered for the overall Hacktoberfest Open-Source AI Challenge: Week 3 — Slow and Steady.
SlothDesk AI uses optional local inference through Ollama to support privacy-conscious file organization without requiring a paid proprietary AI API.
Top comments (0)