DEV Community: Anushka Sharma

SlothDesk 🦥 — Less Mess. More Rest

Anushka Sharma — Sat, 24 Oct 2026 08:28:51 +0000

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 3: Slow and Steady

What I Built

Some tasks aren’t difficult—they’re just easy to keep postponing. Organizing files is one of them.

I built SlothDesk, a digital assistant that helps turn a cluttered folder into an organized download with less repetitive work.

Instead of manually sorting every file, users can:

Upload files or select a folder.

Get category and filename suggestions.

Find exact duplicates and documents with similar filenames.

Review which files to include and how to organize them.

Download the result as an organized ZIP.

View daily summaries of their uploaded workspace.

Each account has its own private uploads. SlothDesk works on uploaded copies, so the original files on your computer stay unchanged.

It’s built for students, developers, and anyone whose idea of organizing files is creating another folder called “New Folder.” 😅

Demo

🦥 Try SlothDesk: https://slothdesk.onrender.com/

Create an account, upload a few sample files, review the suggestions, and download your organized ZIP.

For a quick duplicate-detection demo, upload two identical files with different names. SlothDesk groups exact matches using SHA-256 hashes, so detection goes beyond comparing filenames.

Code

GitHub repository: https://github.com/Anushkasharma101/slothdesk.git

The project includes the frontend, backend, MongoDB models, optional Ollama integration, tests, and deployment instructions.

How I Built It

SlothDesk uses:

React and Vite for the frontend.

Node.js and Express for the backend.

MongoDB Atlas for accounts, upload metadata, review choices, and digest history.

Multer for file uploads.

Archiver for organized ZIP downloads.

Render for hosting the application.

The workflow is straightforward:

Upload → Analyze → Review → Download

Classification starts with file extensions and filename normalization. Exact duplicate detection uses file hashes, while similar-filename groups are flagged separately for manual review.

For AI-assisted naming, I added an Ollama integration that can use a configured open-weight model to suggest cleaner filenames. It sends filenames and categories—not document contents—to the model. Suggestions are validated before being used.

Current AI status: Ollama support is implemented, but an active model in the public deployment has not yet been verified. When no model is configured, or inference is unavailable, SlothDesk uses its built-in rules and labels that behavior.

I also kept the final decision with the user. They can exclude duplicate copies, preserve an uploaded path, or accept the suggested organization before downloading.

Why Does Open Innovation Matter?

The open-model integration gives SlothDesk room to evolve without tying its naming assistant to one proprietary API.

With Ollama, a self-hosted installation can choose a model that fits its hardware, experiment with prompts, and change the naming behavior. When inference runs locally, filename metadata can stay within that installation rather than being sent to a third-party inference service.

That flexibility matters for a tool handling personal digital clutter. Users should be able to understand what the assistant does, where its inputs go, and how its suggestions are produced.

The deterministic parts remain inspectable too: categorization, duplicate detection, path validation, and ZIP generation all have explicit behavior. AI adds suggestions; the user reviews the result.

SlothDesk’s goal is simple: make a small, recurring chore easier to finish—and leave a little more time to take it slow. 🦥

QuoteCraft: Turning Customer Enquiries into Business Quotations

Anushka Sharma — Sat, 17 Oct 2026 17:42:01 +0000

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 2: Side Hustle

What I Built

A customer sends a message:

“Hi! I need two beige wool carpets, size 6×9 feet. Can you share a quotation?”

For a small business owner, answering that question involves several steps: understanding the requirements, checking the catalogue, finding prices, calculating totals, and preparing a document.

I built QuoteCraft to bring those steps into one workspace.

QuoteCraft helps small sellers turn customer enquiries into editable, professional quotations. The first version uses a carpet business as its example, but the workflow could also serve furniture shops, printers, and other businesses with product catalogues.

Users can:

Manage products, specifications, and prices.

Store customer details.

Paste an enquiry and request a quotation draft.

Review matched products and edit quantities or rates.

Add custom line items.

Configure tax, payment terms, and delivery estimates.

Save quotations and track their status.

Export a quotation through the browser’s Save as PDF option.

Who would pay for it? A small business owner who prepares similar quotations repeatedly and wants to reduce manual entry while keeping control over the final price.

My initial business hypothesis is a paid setup and customization service, with ongoing support as an option. I haven’t validated willingness to pay yet.

Demo(frontend)

Demo link or video: https://quotecraft-rose-nine.vercel.app/

A sample walkthrough:

Open the product catalogue.

Paste a customer enquiry.

Generate a draft.

Review the products, quantities, and prices.

Save the quotation and export it as a PDF.

The app includes two distinct drafting modes:

Local AI mode: calls an open-weight model through Ollama.

Demo mode: uses simple keyword matching to test the application without a model.

Demo mode is not AI inference. A deployed frontend and backend still need access to an Ollama service for AI mode to work.

Code

The project repository: https://github.com/Anushkasharma101/quotecraft.git

The code is separated into frontend and backend folders so each can be developed and deployed independently.

How I Built It

QuoteCraft uses:

Frontend: HTML, CSS, JavaScript, and Vite.

Backend: Node.js with a JSON API.

Database: MongoDB for products, customers, quotations, and settings.

AI integration: Ollama, configured to use the open-weight qwen2.5:3b model.

The AI integration handles a specific task: interpreting an enquiry and suggesting catalogue product IDs and quantities.

The backend sends the enquiry and catalogue to Ollama and requests a structured JSON response. It then checks the returned product IDs and retrieves prices from the database.

The model interprets the message; application code calculates the money.

Subtotal, tax, and total are calculated by the backend. Users review the proposed items and can correct them before saving.

Saved quotations also contain snapshots of customer details and line items. Changing a catalogue price later does not automatically rewrite an earlier quotation.

One important lesson was keeping AI integration separate from deployment. Hosting the web application does not automatically host a local model. That led me to keep demo mode clearly labelled rather than silently falling back when AI is unavailable.

Why Does Open Innovation Matter?

For a small business tool, the choice of AI affects both operating costs and how much control the owner has.

Control over inference: With Ollama, the model can run on hardware controlled by the operator. The enquiry does not need to be sent to a closed model API for interpretation. Database privacy still depends on where MongoDB is hosted.

Flexible model choice: The integration is not tied to one model provider. I can experiment with another compatible model and evaluate whether it handles product specifications or customer language better.

A different cost structure: Self-hosting avoids a per-request model API charge. It still has hardware, electricity, or hosting costs, so I wouldn’t describe inference as free.

Room for niche improvements: A quotation assistant needs to understand a business’s catalogue, terminology, and missing requirements. An open approach gives me room to change prompts, validation, and eventually model adaptation around that specific workflow.

I haven’t benchmarked QuoteCraft against a closed API, so I’m not claiming better accuracy or speed. The practical benefit is the ability to build and inspect a quotation workflow around a model I can run and replace.

For me, QuoteCraft is a small, focused experiment: can an open-weight model help a seller spend less time preparing a quote while leaving the business decisions in their hands?

# BreakTrail: Less Scrolling, More Outside 🌿

Anushka Sharma — Sun, 11 Oct 2026 11:04:16 +0000

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 1: Touch Grass.

What I Built

As a developer, I spend much of my day switching between code, meetings, and browser tabs. Even when I take a break, I often end up scrolling on another screen.

I built BreakTrail to turn those small gaps in the day into outdoor moments.

BreakTrail helps users plan a short outdoor break based on their available time, starting area, mood, and budget. Instead of presenting an endless list of recommendations, it offers a practical plan: somewhere to go, something to do, and a reason it fits.

Users can:

Plan a break around moods such as Quiet, Nature, Movement, or Discovery.
Generate an AI activity from a shortlist of suitable places.
Save plans and export them as JSON files or visual cards.
Use an outdoor timer with pause and resume controls.
Record how they felt afterward and what they noticed.
Access cached saved plans and the timer offline.

It is designed for developers, students, remote workers, and anyone who wants a small reason to step outside.

The current prototype includes an illustrative Jaipur place catalogue. It does not provide live navigation, weather, or verified opening hours.

Demo

🌐 Website: https://breaktrail-nu.vercel.app/

A suggested walkthrough:

Choose your area, mood, budget, and available time.
Generate a plan or explore demo mode.
Save the plan and open its outdoor card.
Start the timer.
Add a reflection after your break.

Saved plans and the timer can work offline after the production app is cached. New plans through the hosted AI service require internet.

Code

The source code is available on GitHub:

https://github.com/Anushkasharma101/breaktrail.git

The repository includes the React frontend, Express backend, AI adapters, tests, and deployment configuration.

How I Built It

BreakTrail uses React, TypeScript, Vite, and Tailwind CSS for the frontend. The backend uses Node.js, Express, SQLite, and Zod.

For local inference, I integrated Qwen3 through Ollama, using the qwen3:0.6b model. The lightweight model provides a practical starting point for running the planner locally.

The planning pipeline combines application rules with model-generated text:

The backend filters catalogue entries by area, mood, budget, and available time.
The model receives only the eligible shortlist.
It selects a place and suggests an activity and explanation.
The backend validates the response before building and saving the plan.

This keeps place selection grounded in the catalogue rather than asking the model to invent destinations. Walking estimates are included in the time budget so the plan leaves time for the outdoor activity.

I also added an OpenRouter adapter for free hosted inference. This lets the deployed backend request AI output without requiring visitors to install Ollama. The openrouter/free router can select different available free models, so the hosted version does not guarantee Qwen.

The PWA caches the interface for offline access, while browser storage keeps saved plans and reflections. An offline queue retries backend sync when a connection becomes available.

For public deployment, I added private browser tokens that separate each visitor’s backend records. API keys stay on the server.

Why Does Open Innovation Matter?

Open-weight models make it possible to explore a version of BreakTrail that runs locally, without depending entirely on a hosted inference service.

With Qwen and Ollama, developers can inspect the integration, change the prompts, experiment with model sizes, and run inference on their own hardware. Once the necessary software and model are downloaded, the local setup can generate plans without an internet connection.

A closed hosted API could provide useful recommendations, but it would not give this project the same option to run that model locally.

Sharing the application code also gives others a starting point for adapting the catalogue to their city, improving accessibility information, or building better outdoor activities.

For me, open innovation means making a small idea easier for someone else to understand, change, and build upon.

My Agent Session

I used an AI coding assistant to help inspect the project, prepare deployment changes, add visitor data isolation, and integrate the hosted AI adapter.

The collaboration also helped me troubleshoot practical deployment problems: a backend accidentally tracked as a nested Git repository, changing deployment URLs, CORS configuration, missing device authentication, and AI response validation.

I tested the deployed app and used browser network responses to identify failures. Those checks helped distinguish a successful deployment from a working end-to-end AI flow.

MoveMate: Helping My Friend Move Without Missing Work

Anushka Sharma — Sun, 04 Oct 2026 12:08:26 +0000

This is a submission for the Hacktoberfest Weekend Challenge: Build for a Friend

What I Built

I built MoveMate, a moving coordination platform, for Arnav Kulshrestha, who **The idea for MoveMate came from a friend who lives in Pune and was shifting from one flat to another. His moving day coincided with a critical deployment at work. He needed to attend meetings and stay available to his team while also coordinating the vehicle, workers, packing and moving his belongings.

He found himself juggling two responsibilities that both needed his attention. Taking time off was difficult, but managing the move alongside work was stressful.

This is a common challenge for working professionals in metro cities. I built MoveMate to help him coordinate the move through a dedicated coordinator, review a moving plan and follow progress in one place, so he could stay informed while focusing on his work.**.

Moving involves more than booking a truck. Someone needs to arrange workers, explain packing requirements, coordinate access and follow up throughout the day. MoveMate brings those tasks into one place, helping my friend plan a move around their work schedule.

The application has three portals:

Customer: Create a booking, enter inventory and unavailable hours, nominate access delegates, review an AI-generated moving plan, accept quotes and follow progress.
Coordinator: Apply to join, receive assigned jobs after approval, manage checklists, upload evidence and record moving progress.
Admin: Review coordinator applications, manage vehicles and workers, assign resources, prepare itemised quotes and handle incidents.

Additional costs require a separate customer decision. The customer confirms delivery before the booking is completed.

Demo

Live application: https://frontend-silk-three-78.vercel.app

The demo follows a complete booking journey: customer registration → moving details and inventory → plan review → submission → admin assignment and quote → customer acceptance → coordinator progress → delivery confirmation.

Feedback from my friend: “After trying MoveMate, I liked having the moving plan, assigned team and progress updates in one place. Being able to enter my unavailable hours and nominate someone to provide access was especially useful because I can’t always step away from work.

The quote review and separate approval for extra costs also made the process clearer. I’d want to try it during an actual move, but the booking experience felt organised and addressed the difficulties I faced while shifting in Pune.”

Code

GitHub repository: https://github.com/Anushkasharma101/movemate.git

The project includes the React frontend, Express backend, API documentation and automated tests.

How I Built It

I used React with Vite for the frontend, Node.js and Express for the backend, and MongoDB Atlas for application data.

The open-weight AI model at the centre of the planning feature is Qwen3-30B-A3B-Instruct-2507, accessed through Backboard.io.

During booking, the backend sends relevant inventory and moving constraints to Qwen. It validates the response and displays an editable draft that the customer reviews before submission. This turns the AI output into a practical part of the moving workflow.

AI does not assign providers, confirm availability, set prices or approve extra charges. Those actions remain controlled by people and backend permissions.

The backend is deployed on Render, and the frontend is deployed on Vercel. Backboard credentials stay on the backend. Server-side access checks keep customer, coordinator and admin operations protected.

Why Does Open Innovation Matter?

Using an open-weight model gives me more control over how MoveMate’s planning feature can develop. Beyond changing prompts, I can explore self-hosting or task-specific fine-tuning in a future version—options that a closed model’s weights would not provide.

The backend model setting is configurable, making it easier to experiment with different models without rebuilding the customer interface.

For this weekend’s prototype, hosted Qwen through Backboard let me focus on the moving workflow without setting up local inference. The current application requires internet access and uses a hosted AI service; it does not claim to run offline or cost nothing.

Customers can also review and save a manual plan when AI is unavailable. The goal is useful assistance that people can inspect and edit.

Prize Categories

Best Use of MongoDB Atlas — Atlas provides the database behind MoveMate’s application and open-weight AI planning workflow.

MoveMate also integrates Backboard.io for hosted Qwen access and uses Render for backend hosting.

DEV Community: Anushka Sharma

SlothDesk 🦥 — Less Mess. More Rest

QuoteCraft: Turning Customer Enquiries into Business Quotations

# BreakTrail: Less Scrolling, More Outside 🌿

What I Built

Demo

Code

How I Built It

Why Does Open Innovation Matter?

My Agent Session

**MoveMate: Helping My Friend Move Without Missing Work**

What I Built

Demo

Code

How I Built It

Why Does Open Innovation Matter?

Prize Categories

MoveMate: Helping My Friend Move Without Missing Work