DEV Community: Himanshu Gupta

LazyDesk AI — Your Local AI Productivity Assistant | Work Less, Live More

Himanshu Gupta — Mon, 26 Oct 2026 04:33:40 +0000

LazyDesk AI: Your Local AI Productivity Assistant 🚀

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

What I Built

I built LazyDesk AI, a local-first AI productivity assistant that helps users spend less time on repetitive tasks and more time on meaningful work.

Students, developers, and busy individuals often spend time summarizing notes, creating task lists, organizing filenames, and planning their day. LazyDesk AI brings these workflows together in one simple dashboard.

Features

📝 Smart Summary: Convert long text into concise, easy-to-understand summaries.
✅ Auto Task Maker: Turn notes and ideas into actionable checklists.
📁 File Organizer: Get cleaner filename suggestions and export a rename plan.
☀️ Morning Brief: Generate a productivity brief based on pending tasks.
🔒 Local AI: Run compatible AI models locally using Ollama.
🌙 Dark Mode: Switch between light and dark themes.

The idea is simple: let AI help with repetitive work so people can focus on what matters.

Demo

🌐 Live Website: https://himanshugupta4105.github.io/LazyDesk-AI/#dashboard

The frontend is hosted on GitHub Pages. The AI features require Ollama to be installed and running locally on the user's computer.

Code

💻 GitHub Repository: https://github.com/Himanshugupta4105/LazyDesk-AI

The project uses three main frontend files:

index.html — Dashboard structure
style.css — Styling and responsive layout
script.js — UI interactions and AI integration

I chose a lightweight frontend so that the project remains easy to understand, modify, and extend.

How I Built It

I built LazyDesk AI using HTML, CSS, JavaScript, and Ollama.

The project is designed to use Qwen2.5 3B, an open-weight AI model that can run locally through Ollama.

Here is how the AI workflow operates:

The user enters text or selects a productivity feature.
JavaScript prepares a prompt for the selected task.
The application sends the prompt to Ollama's local API.
The AI model generates a response.
The dashboard displays the result.

For file organization, the application suggests cleaner filenames and creates an exportable rename plan instead of automatically changing files.

One important detail: GitHub Pages hosts the frontend, not the AI model. Users need a local Ollama installation, a downloaded model, and suitable browser connection settings to use the AI features.

Why Does Open Innovation Matter?

Open innovation makes AI experimentation more accessible to students, independent developers, and open-source communities.

Using Ollama and an open-weight model allows LazyDesk AI to explore useful AI workflows without depending on a paid, hosted AI API.

This approach offers several benefits:

Accessibility: Developers can experiment with local AI without per-request hosted API fees.
Privacy-conscious design: Prompts can remain on the user's computer when local inference is configured correctly.
Flexibility: Users can choose compatible models and customise prompts.
Transparency: Developers can inspect and improve the frontend code.
Community collaboration: Other developers can extend the project and contribute new features.

Open innovation is about more than making software available. It gives people the freedom to learn, experiment, adapt, and build on existing technology.

My Agent Session

I developed the project iteratively, working on the frontend, local AI integration, and documentation.

I do not yet have a verified DevRelay agent-session link to share, so I have left out an unverified recording. I hope to include a session record in a future update.

The Lazy Payoff

LazyDesk AI is designed to reduce the small, repetitive tasks that interrupt focused work.

Instead of manually extracting key points from notes, users can generate a summary. Instead of writing every task from scratch, they can generate a checklist. The file organizer also helps standardize filenames.

I have not yet conducted a controlled test to measure the actual time saved, so I will not claim a specific number.

My next step is to test these workflows with users, measure the time required with and without the assistant, and improve the experience based on their feedback.

Prize Categories

Project focus: Open-Source AI and Local Inference.

LazyDesk AI explores how an open-weight model running locally can support everyday productivity through a lightweight web application.

I will select the applicable partner prize categories according to the official challenge rules and the project's eligibility.

Final Thoughts

LazyDesk AI started with one simple question:

What if AI could handle repetitive tasks so we could focus on what really matters?

This project is my step toward exploring that idea using open-source tools, local inference, and familiar web technologies.

It is a work in progress, and I look forward to learning from the community, receiving feedback, and making it more useful.

Thanks for reading! ⭐

Built with HTML, CSS, JavaScript, and Ollama.

LazyDesk AI — Work less, live more.

MarketMate AI: Turn Product Details into Listings That Sell.

Himanshu Gupta — Thu, 15 Oct 2026 16:28:01 +0000

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

MarketMate AI: Turn Product Details into Listings That Sell 🚀

What I Built

MarketMate AI is an AI-powered product listing generator designed for small business owners, online sellers, freelancers, and entrepreneurs.

Creating attractive product listings takes time. Sellers need to write product descriptions, social media captions, and hashtags that help their products stand out. Not everyone has the time, writing skills, or budget to hire a professional copywriter.

MarketMate AI aims to make this process easier.

Users can enter their product details, choose a category and writing tone, and generate marketing content with AI.

Key Features

📝 Generate product titles and descriptions.
📱 Create social media captions.
#️⃣ Generate relevant hashtags.
🎯 Customize the tone and language of the content.
📋 Copy generated content for use in product listings and marketing.
🔓 Use open-source AI through local inference with Ollama.

Who Would Pay for It?

Potential customers include small business owners, Instagram sellers, resellers, freelancers, and people starting an online side hustle.

A future paid version could offer bulk listing generation, saved templates, multiple product variations, and additional marketing tools.

My goal is to make AI-powered marketing more accessible to people who cannot afford expensive marketing software.

Demo

🌐 Try MarketMate AI:

https://himanshugupta4105.github.io/MarketMate-AI/

Code

💻 GitHub Repository:

https://github.com/Himanshugupta4105/MarketMate-AI

The project uses HTML, CSS, and JavaScript for the interface and application logic, with a local Ollama model intended to generate marketing content.

Explore the repository, try the demo, and share your feedback!

How I Built It

I built MarketMate AI using familiar web technologies and an open-source AI approach.

Technology stack:

HTML — structures the application and its forms.
CSS — creates the responsive user interface.
JavaScript — handles user input, API requests, and generated content.
Ollama — provides a way to run an open-weight language model locally.
Llama 3.2 — the model configured in the project for generating marketing content.

The intended local AI workflow is straightforward:

The user enters product information.
The application sends a structured prompt to the local AI model.
The model generates marketing content.
The application displays the results so the user can copy and use them.

A local Ollama setup is required for the local inference workflow; the GitHub Pages demo alone does not host an Ollama model.

Why Does Open Innovation Matter?

Open innovation makes AI development more accessible to independent developers and small businesses.

For MarketMate AI, using open-weight models creates several opportunities:

More control: Developers can choose and change models based on their needs.
Privacy potential: With properly configured local inference, product information can be processed on the user's own machine.
Cost flexibility: Local inference can reduce dependence on per-request fees from hosted AI APIs, although hardware and electricity still have costs.
Experimentation: Developers can test different prompts and models without rebuilding the entire application around a single closed provider.
Accessibility: Small businesses can potentially access useful AI tools without paying for an expensive subscription from day one.

Open-source and open-weight AI are not automatically the same thing, and model licenses vary. Developers should check the selected model's license and deployment requirements.

For me, the biggest benefit is the freedom to experiment, learn, and build a useful product without depending entirely on a closed AI API.

My Agent Session

I used AI-assisted development to help plan and build the project.

I haven't included a DevRelay session link in this submission. If I publish a saved agent session, I'll add it here.

Prize Categories

My main entry is the Side Hustle challenge: building a practical product around open-source AI with potential real-world value.

MarketMate AI: Turn Product Details into Listings That Sell.

Himanshu Gupta — Thu, 15 Oct 2026 15:04:44 +0000

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

MarketMate AI: Turn Product Details into Listings That Sell 🚀

What I Built

MarketMate AI is an AI-powered product listing generator designed for small business owners, online sellers, freelancers, and entrepreneurs.

MarketMate AI aims to make this process easier.

Users can enter their product details, choose a category and writing tone, and generate marketing content with AI.

Key Features

📝 Generate product titles and descriptions.
📱 Create social media captions.
#️⃣ Generate relevant hashtags.
🎯 Customize the tone and language of the content.
📋 Copy generated content for use in product listings and marketing.
🔓 Use open-source AI through local inference with Ollama.

Who Would Pay for It?

Potential customers include small business owners, Instagram sellers, resellers, freelancers, and people starting an online side hustle.

A future paid version could offer bulk listing generation, saved templates, multiple product variations, and additional marketing tools.

My goal is to make AI-powered marketing more accessible to people who cannot afford expensive marketing software.

Demo

🌐 Try MarketMate AI:

https://himanshugupta4105.github.io/MarketMate-AI/

Code

💻 GitHub Repository:

https://github.com/Himanshugupta4105/MarketMate-AI

The project uses HTML, CSS, and JavaScript for the interface and application logic, with a local Ollama model intended to generate marketing content.

Explore the repository, try the demo, and share your feedback!

How I Built It

I built MarketMate AI using familiar web technologies and an open-source AI approach.

Technology stack:

HTML — structures the application and its forms.
CSS — creates the responsive user interface.
JavaScript — handles user input, API requests, and generated content.
Ollama — provides a way to run an open-weight language model locally.
Llama 3.2 — the model configured in the project for generating marketing content.

The intended local AI workflow is straightforward:

The user enters product information.
The application sends a structured prompt to the local AI model.
The model generates marketing content.
The application displays the results so the user can copy and use them.

A local Ollama setup is required for the local inference workflow; the GitHub Pages demo alone does not host an Ollama model.

Why Does Open Innovation Matter?

Open innovation makes AI development more accessible to independent developers and small businesses.

For MarketMate AI, using open-weight models creates several opportunities:

More control: Developers can choose and change models based on their needs.
Privacy potential: With properly configured local inference, product information can be processed on the user's own machine.
Cost flexibility: Local inference can reduce dependence on per-request fees from hosted AI APIs, although hardware and electricity still have costs.
Experimentation: Developers can test different prompts and models without rebuilding the entire application around a single closed provider.
Accessibility: Small businesses can potentially access useful AI tools without paying for an expensive subscription from day one.

Open-source and open-weight AI are not automatically the same thing, and model licenses vary. Developers should check the selected model's license and deployment requirements.

For me, the biggest benefit is the freedom to experiment, learn, and build a useful product without depending entirely on a closed AI API.

My Agent Session

I used AI-assisted development to help plan and build the project.

I haven't included a DevRelay session link in this submission. If I publish a saved agent session, I'll add it here.

Prize Categories

My main entry is the Side Hustle challenge: building a practical product around open-source AI with potential real-world value.

🌿 GreenQuest AI: Turn AI Time Into Outside Time

Himanshu Gupta — Wed, 07 Oct 2026 12:59:28 +0000

🌿 GreenQuest AI: Turn AI Time Into Outside Time

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

What I Built

I built GreenQuest AI, an open-source AI-powered web application designed to help people spend less time on screens and more time outdoors.

The idea is simple:

AI plans the adventure → You close the screen → You go outside. 🌱

GreenQuest AI asks the user for:

⏱️ Available time
😊 Mood
🌳 Outdoor location

It then uses a local open-weight AI model through Ollama to generate a personalized outdoor quest with 5 simple activities.

For example, a user can choose:

30 minutes + Relaxing + Park

GreenQuest AI might create a short nature exploration quest involving observing the surroundings, walking, noticing sounds, and finding interesting natural details.

The project is designed for students, developers, remote workers, and anyone who feels they are spending too much time in front of a screen.

The goal isn't to keep users inside the app.

The goal is to get them out of it.

Demo

🌐 Live Website:
https://himanshugupta4105.github.io/GreenQuest-AI/

You can try GreenQuest AI directly in your browser.

Note: The AI-powered generation works with Ollama running locally. The project also includes an offline fallback so users can still experience the core functionality.

Code

💻 GitHub Repository:
https://github.com/Himanshugupta4105/GreenQuest-AI

The project is built using:

HTML
CSS
JavaScript
Ollama
Open-weight AI model
Local browser storage

There is no traditional backend or database required.

How I Built It

The main idea behind GreenQuest AI was to use open-source AI as a bridge between digital technology and real-world experiences.

Instead of building another AI chatbot that keeps users talking to the screen, I wanted AI to do the planning and then encourage the user to leave the screen.

AI Stack

GreenQuest AI uses:

Ollama for local AI inference.

The default model is:

llama3.2

The application sends a structured prompt to the local Ollama API:

http://localhost:11434/api/generate

The AI generates a structured outdoor quest containing:

Quest title
Introduction
5 outdoor tasks
Duration
Nature/safety tip

The application then displays the generated quest as an interactive checklist.

Project Architecture

User
  ↓
GreenQuest AI Web Interface
  ↓
JavaScript
  ↓
Ollama Local API
  ↓
Open-Weight AI Model
  ↓
Personalized Outdoor Quest
  ↓
User Leaves the Screen
  ↓
Real-World Experience 🌱

I also added an offline fallback system. If Ollama is unavailable, GreenQuest AI can still generate predefined outdoor quests.

The application stores progress such as completed quests, outdoor minutes, and streaks using browser localStorage.

Why Does Open Innovation Matter?

Open innovation was important for GreenQuest AI because the project is built around the idea of local and accessible AI.

Using an open-weight model with Ollama provides several advantages.

🔒 Privacy

The AI can run locally on the user's computer.

Personal inputs don't have to be sent to a third-party cloud AI service.

🧠 Model Freedom

Because the application uses Ollama, developers can experiment with different open-weight models instead of being locked into one closed API.

💰 Lower Cost

Local inference means there is no per-request cloud AI API cost for the application.

🛠️ Customization

Developers can modify the prompts, change the model, experiment with different AI behaviors, and improve the project according to their needs.

🌍 Open Collaboration

Anyone can inspect the code, understand how the AI integration works, suggest improvements, and contribute new ideas.

For GreenQuest AI, open innovation isn't just a technical choice.

It supports the project's philosophy:

AI should help us experience more of the real world, not keep us trapped inside a screen.

My Agent Session

I used AI-assisted development while building GreenQuest AI to help with project planning, UI ideas, JavaScript logic, and AI integration.

Agent session:

Prize Categories

I am submitting GreenQuest AI for the Hacktoberfest Open-Source AI Challenge Week 1: Touch Grass.

The project focuses on:

🌿 Getting people outdoors
🤖 Open-source/local AI
🔓 Open innovation
💻 Developer experimentation
🌎 Real-world impact

What Makes GreenQuest AI Different?

There are already many AI applications designed to keep people engaged with screens.

GreenQuest AI takes the opposite approach.

The best outcome isn't another minute spent using the application.

The best outcome is:

The user closes the browser and goes outside. 🌱

The AI is intentionally only a small part of the experience.

It plans the activity, and then the real experience happens away from the screen.

Future Improvements

Some features I would like to add in future versions include:

🗺️ Outdoor location suggestions
🌤️ Weather-aware quests
🐦 Nature and wildlife recognition
🏆 More gamification
📱 Progressive Web App support
📊 More detailed outdoor statistics
🌱 Community-created quests
🤖 Support for more open-weight AI models
🌍 Multi-language support

Safety & Nature Guidelines

GreenQuest AI is designed around simple and responsible outdoor activities.

Generated quests should:

Avoid dangerous activities
Avoid disturbing wildlife
Avoid damaging plants or nature
Avoid requiring expensive equipment
Encourage users to stay in familiar and appropriate outdoor areas
Keep activities realistic and accessible

🤝 Open Source

GreenQuest AI is open to improvements and contributions.

Developers can contribute by:

Improving the UI
Adding new quest types
Improving AI prompts
Supporting additional models
Adding accessibility features
Improving offline functionality
Adding new languages
Fixing bugs
Suggesting new outdoor experiences

🌿 Final Thought

Technology doesn't always have to keep us online.

Sometimes the best thing AI can do is help us put the phone down and experience the world around us.

AI plans it.
You experience it.
Less screen. More world. 🌱

🔗 Links

Live Demo:
https://himanshugupta4105.github.io/GreenQuest-AI/

GitHub:
https://github.com/Himanshugupta4105/GreenQuest-AI

🌿 TrailMate AI — Less Screen, More World with Open-Source AI

Himanshu Gupta — Wed, 07 Oct 2026 05:46:18 +0000

🌿 TrailMate AI — Less Screen, More World with Open-Source AI

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

What I Built

TrailMate AI is a nature-focused web application designed to help people spend less time looking at screens and more time exploring the real world.

The idea is simple:

Use technology to help people leave technology. 🌱

TrailMate AI gives users small outdoor missions based on activities such as:

🚶 Walking
🥾 Hiking
🐦 Bird watching
🌿 Nature exploration

Instead of continuously scrolling on a phone, users get a simple mission that encourages them to go outside, observe their surroundings, and discover something new.

What TrailMate AI Does

The application includes:

🌱 Outdoor mission generator
⏱️ 10-minute outdoor activity timer
📸 Nature discovery/photo upload
🤖 AI discovery interface
📊 Outdoor progress tracking
🔥 Streak tracking
💾 Local progress storage
📱 Responsive design for desktop and mobile
🌐 Online/offline status indicator

For example, a user can open TrailMate AI and receive a mission such as:

"Take a short walk and notice three different things in nature."

The goal is not to keep the user inside the app.

The goal is to get the user outside and then close the app.

Who Is It For?

TrailMate AI is designed especially for:

Students
Developers
People who spend long hours on computers
People who want to explore nature
Anyone who wants a simple reason to take a break from screens

Demo

🚀 Live Website

Try TrailMate AI here:

https://himanshugupta4105.github.io/TrailMate-AI/

The website is hosted using GitHub Pages.

Code

💻 GitHub Repository

The complete project is available on GitHub:

https://github.com/Himanshugupta4105/TrailMate-AI

Project Structure

The project is intentionally simple and beginner-friendly:

TrailMate-AI/
│
├── index.html
├── style.css
├── script.js
└── README.md

Technologies Used

HTML5
CSS3
JavaScript
Browser LocalStorage
GitHub Pages

No frontend framework is required.

The project can be opened directly in a browser.

How I Built It

I wanted to build something that uses technology for a different purpose: helping people spend less time with technology.

I built the first version using only three files:

HTML → Structure
CSS → Design
JavaScript → Functionality

1. HTML

HTML creates the structure of the application.

It contains:

Navigation
Hero section
Mission section
Timer
Discovery section
Progress section
Footer

2. CSS

CSS gives TrailMate AI its nature-inspired interface.

I designed the interface around:

Green/nature colors
Rounded cards
Clean typography
Responsive layouts
Nature-inspired visual elements
Mobile-friendly design

3. JavaScript

JavaScript powers the interactive features.

It handles:

Generating missions
Selecting activities
Completing missions
Outdoor timer
Image upload
Discovery interface
Progress tracking
Streaks
LocalStorage
Online/offline detection
Notifications

AI Approach

The project is designed around an open-source AI architecture.

The current public prototype contains a simulated AI discovery layer so that the frontend can be demonstrated without requiring users to expose an API key in browser-side JavaScript.

The next AI implementation is intended to use an open-weight model running locally, rather than depending entirely on a closed cloud API.

The planned flow is:

User goes outside
       ↓
Takes a photo
       ↓
TrailMate AI
       ↓
Local/Open-Weight AI Model
       ↓
Identifies the discovery
       ↓
User learns about it
       ↓
User continues exploring

This approach also makes it possible to move toward offline and privacy-friendly AI.

Why Does Open Innovation Matter?

Open innovation is important for TrailMate AI because the project is about making AI more accessible, transparent, and useful outside traditional cloud-only applications.

A closed AI API can be useful, but it can also introduce:

API costs
Internet dependency
Rate limits
Privacy concerns
Vendor dependency
Limited control over the model

With open-weight and open-source AI, developers can have more control over how AI is used.

For TrailMate AI, this could make it possible to eventually run AI locally on a user's device.

That means:

Camera
   ↓
Local AI
   ↓
Nature Identification
   ↓
No unnecessary cloud upload

What Open AI Makes Possible

Open innovation can help TrailMate AI move toward:

🔒 Better privacy
📡 Offline functionality
💰 Lower long-term AI costs
🧠 More control over models
🛠️ Community improvements
🔄 Ability to experiment with different models
🌍 More accessible AI development

Most importantly, open-source AI allows developers to build, experiment, modify, and learn instead of treating AI as a black box.

My Agent Session

This section is optional.

I have not included an Agent Session link yet.

If I create a DevRelay agent session for TrailMate AI, I will add it here:

[Add DevRelay Agent Session link here]

Prize Categories

I am submitting TrailMate AI for the following challenge category:

🌱 Hacktoberfest Open-Source AI Challenge — Week 1: Touch Grass

The project focuses on the core idea of the challenge:

Use open-source AI and technology to encourage people to spend more time in the real world.

TrailMate AI is designed around a simple philosophy:

Less screen. More world. 🌎

Future Improvements

TrailMate AI is currently an early prototype, but there are several features I would like to add.

🤖 Real AI Nature Identification

Replace the simulated discovery system with a real open-weight computer vision model.

📴 Better Offline AI

Allow nature identification without requiring an internet connection.

🗺️ Outdoor Exploration

Add safe exploration areas and routes.

🐦 Bird Identification

Use AI to identify birds from photos or potentially their calls.

🌿 Plant Identification

Help users learn about plants they discover.

🏆 Community Challenges

Allow users to participate in outdoor challenges with friends.

📈 Better Progress Tracking

Track:

Outdoor time
Missions completed
Discoveries
Streaks
Favorite activities

🔐 Privacy-First AI

Keep image processing local wherever possible instead of automatically sending personal images to a remote server.