DEV Community: Muskan

Receipt Pilot

Muskan — Sun, 25 Oct 2026 16:01:57 +0000

ReceiptPilot — Your Smarter Expense Tracker 🧾

Meet ReceiptPilot, a lightweight expense-tracking dashboard designed to make managing everyday expenses easier.

✨ What it does:

📸 Upload receipt images and extract text using OCR.
✏️ Review and edit extracted merchant, date, amount, and category details.
📊 Track total spending and explore category-wise expense insights.
🔎 Search and filter saved expenses.
📥 Export expense records to CSV and back them up as JSON.
🌙 Switch between light and dark themes.

Built with: HTML, CSS, JavaScript, and Tesseract.js OCR.

ReceiptPilot stores expense records locally in the browser, keeping this prototype simple and accessible without requiring a backend.

🚀 Try it out: https://muskankr.github.io/ReceiptPilot/

💻 Source code: https://github.com/Muskankr/ReceiptPilot

Built as a practical side-hustle project to explore how a simple web application can solve an everyday problem.

DEVChallenge #Hacktoberfest #SideHustle #WebDevelopment #JavaScript #BuildInPublic

Receipt pilot week 2 submission

Muskan — Sun, 25 Oct 2026 13:39:27 +0000

ReceiptPilot — Your Smarter Expense Tracker 🧾

Meet ReceiptPilot, a lightweight expense-tracking dashboard designed to make managing everyday expenses easier.

✨ What it does:

📸 Upload receipt images and extract text using OCR.
✏️ Review and edit extracted merchant, date, amount, and category details.
📊 Track total spending and explore category-wise expense insights.
🔎 Search and filter saved expenses.
📥 Export expense records to CSV and back them up as JSON.
🌙 Switch between light and dark themes.

Built with: HTML, CSS, JavaScript, and Tesseract.js OCR.

ReceiptPilot stores expense records locally in the browser, keeping this prototype simple and accessible without requiring a backend.

🚀 Try it out: https://muskankr.github.io/ReceiptPilot/

💻 Source code: https://github.com/Muskankr/ReceiptPilot

Built as a practical side-hustle project to explore how a simple web application can solve an everyday problem.

DEVChallenge #Hacktoberfest #SideHustle #WebDevelopment #JavaScript #BuildInPublic

LazyDesk Week 3 Submission

Muskan — Sun, 25 Oct 2026 06:56:36 +0000

LazyDesk: A Smarter Way to Tidy Up Your Digital Clutter 🦥

Have you ever opened your Downloads folder and found screenshots, documents, archives, and random files all mixed together?

I built LazyDesk to make dealing with digital clutter simpler. Instead of manually inspecting every file, you can use one dashboard to explore, search, categorize, and review your files.

The idea behind LazyDesk is simple: less repetitive work, more time for things that matter.

🚀 What I Built

LazyDesk is a lightweight, browser-based file organization dashboard built with HTML, CSS, and JavaScript.

It includes:

Smart categorization: Organizes files into categories using their extensions and filename keywords.
Search and filters: Quickly find files without scrolling endlessly.
Grid and table views: Explore your files in the layout you prefer.
Duplicate candidates: Identify files with matching names and sizes for manual review.
Export options: Export file information as CSV or JSON.
Activity log: Keep track of actions performed in the dashboard.
Light and dark themes: Choose a comfortable interface.
Optional semantic AI categorization: Uses Transformers.js with an open-source embedding model to explore meaning-based categorization.

🦥 The Lazy Payoff

The goal isn't to add another complicated productivity tool. It's to reduce the little repetitive tasks that consume time and attention.

Instead of searching through a cluttered interface, I can use filters, categories, and search to narrow things down quickly. Exporting metadata also makes it easier to review a collection outside the app.

LazyDesk is designed as a lightweight starting point for a more automated personal file assistant.

🧠 Why Open-Source AI Matters

I explored Transformers.js and the open-source Xenova/all-MiniLM-L6-v2 model for optional semantic categorization.

The appeal of open-source AI is flexibility. Developers can inspect the implementation, experiment with different models, and build features without committing every part of the workflow to a closed AI service.

Browser-based inference can also reduce the need to send file metadata to a remote AI API. The model may need to download before it can be used, so I don't claim that the current AI feature works completely offline.

The core categorization rules work independently of the AI feature, which keeps the dashboard lightweight and usable even when semantic categorization isn't available.

🛠️ Built With

HTML
CSS
JavaScript
Transformers.js
Xenova/all-MiniLM-L6-v2 for optional semantic categorization

🔗 Explore LazyDesk

Source code: https://github.com/Muskankr/LazyDesk
Live demo or demo video: https://lazydesk-swart.vercel.app/

📚 What I Learned

Building LazyDesk reminded me that useful software doesn't always need to be enormous or complicated. A focused interface, thoughtful organization, and an optional open-source AI experiment can be a good foundation for solving an everyday problem.

My next goal is to explore safer, more automated file-management workflows while keeping user control and privacy in mind.

🌍 Why I Support Open Innovation

Open innovation gives developers the freedom to learn from inspectable tools, experiment with different approaches, and build on technology that others can examine and improve.

For me, LazyDesk is a step toward making everyday digital work less tedious—and toward learning how open-source AI can make small tools more useful.

I'd love to hear your feedback and ideas for making LazyDesk better!

devchallenge #hf26challenge