DEV Community: Devansh Kushwaha

ListingLift: Turn Product Details into SEO-Ready Listings with Open-Source AI

Devansh Kushwaha — Thu, 15 Oct 2026 09:23:23 +0000

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

What I Built

Meet ListingLift — Local AI Listing Studio, a privacy-focused AI tool that helps online sellers turn basic product information into polished, marketplace-ready listings.

Writing product listings takes time. Sellers need catchy titles, detailed descriptions, relevant search tags, and compelling calls to action. ListingLift simplifies this process by generating all four from a single product form.

Key Features

SEO-friendly titles to help make product listings more discoverable.
Product descriptions tailored to the details provided.
Search tags to improve keyword coverage.
Calls to action (CTAs) to encourage potential buyers to take the next step.
Copy and export options to make generated content easy to use.
Local-first AI to keep product information on the user's own computer when using a local Ollama instance.

ListingLift is designed for independent sellers, small businesses, and marketplace entrepreneurs who want to spend less time writing and more time building their businesses.

The potential business model is a lightweight paid tool for sellers who regularly create product listings, with privacy and reduced recurring AI API costs as key advantages.

Demo

Try ListingLift live:

https://devanshkushwaha2008.github.io/ListingLift/

Enter your product details, select your marketplace and brand voice, add keywords, and generate your listing.

Note: The deployed interface may require a locally running Ollama instance to generate AI content.

Code

GitHub Repository:

https://github.com/DevanshKushwaha2008/ListingLift

Built with:

HTML for the interface and form structure.
CSS for the responsive, modern design.
JavaScript for application logic and AI integration.
Ollama for running open-weight AI models locally.

Feel free to explore the code, experiment with different models, and contribute improvements!

How I Built It

I built ListingLift using HTML, CSS, JavaScript, and Ollama, with a focus on making local AI useful for a real-world business problem.

Here's how it works:

Collect product information: Users enter the product name, marketplace, brand voice, product details, ideal customer, and SEO keywords.
Prepare the prompt: JavaScript combines the inputs into instructions for the AI model.
Generate the listing: The application sends the prompt to a locally running Ollama instance, using an installed open-weight model such as Llama 3.2.
Display structured results: The generated title, description, search tags, and call to action appear in separate sections.
Export the content: Users can copy individual sections or export the complete listing as a text file.

The project uses a lightweight frontend instead of a large application backend. This keeps the architecture simple and makes the project easier to understand, customize, and extend.

Why Does Open Innovation Matter?

For ListingLift, open innovation makes local, private, and customizable AI possible.

Many AI-powered writing tools depend on hosted APIs, which can introduce recurring usage costs and require sending data to external services.

By building around Ollama and open-weight models, ListingLift takes a different approach:

Privacy: Product details can be processed on the user's own computer rather than sent to a third-party AI API.
Model freedom: Users can experiment with compatible open-weight models instead of depending on one proprietary model provider.
Cost flexibility: After downloading a model, local inference avoids per-request API charges, although hardware and electricity still cost money.
Customization: Developers can modify prompts, adjust the application, and experiment with different models.
Accessibility: The project provides a starting point for small sellers who want to explore AI-assisted workflows without building an entire AI infrastructure from scratch.

Open innovation is not just an extra feature here. It shapes the way ListingLift handles data, gives developers control, and makes the project open to further experimentation.

My Agent Session

I don't have a public DevRelay agent-session link to share yet, so I'm leaving this section out for now.

Prize Categories

Primary category: Open-Source AI / Local AI

ListingLift uses an open-weight AI model through Ollama, with local inference at the center of its listing-generation workflow.

The project also explores practical AI for small businesses, privacy-conscious software, and developer-friendly open-source tools.

Built with curiosity, open-source technology, and the idea that useful AI tools should give users more control over their data.

Thanks for checking out ListingLift!

TouchGrass AI — Turn Screen Time into Green Time with Smart Outdoor Adventures

Devansh Kushwaha — Fri, 09 Oct 2026 11:10:00 +0000

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

What I Built

TouchGrass AI 🌿 — Turn Screen Time into Green Time with Smart Outdoor Adventures

I built TouchGrass AI 🌿, a web-based outdoor activity planner designed to help people spend less time staring at screens and more time exploring the real world.

Sometimes, we want to go outside but don't know what to do. TouchGrass helps solve that problem by suggesting activities based on your mood, available time, surroundings, and interests.

Users can choose whether they want something relaxing, adventurous, educational, or fitness-focused and get a suggested outdoor plan.

Key Features

🌳 Personalized activity suggestions based on mood, time, and environment.
⏱️ Flexible durations, from a quick 15-minute break to a longer outdoor adventure.
🧭 Different activity categories, including relaxation, exploration, learning, and fitness.
🎯 Nature challenges to encourage users to try something new.
📋 Copyable activity plans that users can take with them outdoors.
📊 Progress tracking to encourage users to keep building outdoor habits.
🔒 Privacy-friendly design, with progress stored locally in the browser.
📱 Responsive interface designed for different screen sizes.

The project is for students, remote workers, and anyone who wants a simple nudge to step away from their devices and reconnect with the world around them.

The main idea is simple: technology should help us experience more of the real world, not keep us trapped in front of a screen.

Demo

🌐 Live Website:
https://devanshkushwaha2008.github.io/TouchGrassAI/

Try choosing your mood, selecting how much time you have, and generating an outdoor activity plan!

Code

💻 GitHub Repository:
https://github.com/DevanshKushwaha2008/TouchGrassAI

The project uses HTML, CSS, and JavaScript, with a simple structure that makes it easy for beginners to explore and contribute.

Feel free to check out the code, suggest improvements, or contribute new activity ideas.

How I Built It

I built TouchGrass AI using:

HTML to structure the website.
CSS to create a clean, nature-inspired, responsive interface.
JavaScript to generate activity plans, handle user interactions, copy plans, and track progress.
Local storage to save progress in the user's browser without requiring an account or a database.
GitHub Pages to deploy the website.

For the current version, activity recommendations are generated using predefined templates and rule-based JavaScript logic. It works without installing additional software, configuring an API key, or running a backend server.

Although the current implementation does not yet run an open-weight AI model, the project is designed around the idea of using technology to encourage healthier, offline experiences. Integrating an open-weight model for more flexible activity planning is a potential next step.

Why Does Open Innovation Matter?

Open innovation makes it possible for developers to learn from one another, experiment with new ideas, and build tools that others can inspect, modify, and improve.

For a project like TouchGrass, openness matters in several ways:

Accessibility: A simple web-based implementation makes it easy for people to try the project without installing specialized software.
Transparency: The recommendation logic is visible in the source code, so users and contributors can understand how suggestions are produced.
Community contributions: Developers can add activities, improve accessibility, refine the interface, and adapt the project for different communities.
Privacy: The current version can operate locally in the browser, without sending users' activity preferences to an external AI service.
Room to experiment: An open codebase gives contributors a starting point for exploring open-weight models and more advanced recommendation systems.

I believe open-source development is valuable because it lets people build on existing ideas instead of starting from scratch. It also gives beginners like me an opportunity to learn by creating something useful and sharing it with others.

My goal with TouchGrass is to make technology a bridge to the outdoors rather than another reason to stay indoors.

My Agent Session

No agent session is linked for this submission.

Prize Categories

This is a project submission for the Touch Grass challenge.

The current implementation is a rule-based web application and does not yet include an open-weight AI model or agent framework. I would integrate and document one before claiming eligibility for categories that specifically require open-source AI at the core.

Thanks for checking out TouchGrass AI! 🌿

If you have ideas for new outdoor activities or ways to make the experience better, contributions and feedback are welcome.

Build with technology. Reconnect with nature. Touch grass.