DEV Community: Sanjay Kumar Mahto

ItemPilot: Turning Product Photos into Buyer-Focused eBay Listings with Local AI 🚀

Sanjay Kumar Mahto — Mon, 19 Oct 2026 06:30:18 +0000

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

What I Built

I built ItemPilot, a local AI-powered eBay Listing Assistant that helps online sellers turn product photos and product details into compelling, editable listing drafts.

Writing product listings can be repetitive. Sellers need to come up with a search-friendly title, describe the product, highlight its useful features, and make the listing appealing to potential buyers.

ItemPilot aims to make that process easier.

Upload a product photo, provide the details you know, and let Gemma 3 4B generate a listing draft containing:

Search-friendly eBay titles with an 80-character limit
Buyer-focused product descriptions
Key features and suggested item specifics
Relevant search keywords
Questions about missing information that the seller should verify

You can review and edit the generated content, then copy or download it before using it on eBay.

Who could pay for it?

The intended users are small eBay sellers, casual resellers, and people who regularly list second-hand electronics, clothing, and other products. For sellers handling multiple listings, saving time on product copy could be valuable.

I haven't validated willingness to pay with a real seller yet. That's an important next step for the project.

Demo

Code

The repository contains the application, prompts, local AI integration, validation logic, tests, and setup instructions.

sanjay067-cse / ItemPilot

A local AI-powered eBay listing assistant using Gemma 3 and Ollama

ItemPilot 🚀

Turn product photos into buyer-focused eBay listings with local AI.

ItemPilot is a local-first AI application that helps online sellers turn product photos and product details into compelling, editable eBay listing drafts using Gemma 3 4B and Ollama.

Instead of starting from a blank page, sellers can upload a product photo, provide the details they know, and generate a polished listing they can review and refine before publishing.

✨ Features

📸 Image-based listing generation — Analyze an uploaded product photo using Gemma 3 4B's vision capabilities.
✍️ Buyer-focused copywriting — Generate engaging, natural product descriptions that highlight the benefits of confirmed features.
🔎 Search-friendly titles — Create relevant eBay listing titles with an 80-character limit.
📋 Structured listing details — Generate product descriptions, key features, item specifics, and relevant search keywords.
🛡️ Accuracy safeguards — Identify missing information and prompt sellers to confirm important details rather than…

View on GitHub

How I Built It

I built ItemPilot using Python, Streamlit, and Gemma 3 4B running locally through Ollama.

Here's how the application works:

The seller uploads a product photo and enters any known product details.
The application sends the image and supplied information to the local Gemma model.
Gemma analyzes the product image and generates listing content.
Python and Pydantic help validate the structured output, including the title-length constraint.
The seller reviews, edits, copies, or downloads the generated listing

My stack:

Python : application logic
Streamlit : user interface
Gemma 3 4B : image understanding and listing generation
Ollama : local model runtime
Pydantic : structured data validation
Pillow : image validation and processing
Pytest : automated testing

I also added safeguards against unsupported product claims. A photograph cannot reliably establish every product specification, so sellers need to confirm important details before publishing.

The current automated test suite passes 36 tests covering configuration, image validation, schemas, listing-service behaviour, title validation, and exports.

Why Does Open Innovation Matter?

I wanted to explore what becomes possible when an AI application is built around a locally running open-weight model rather than a hosted AI API.

With Gemma 3 4B running through Ollama, ItemPilot can perform its core AI generation workflow on the user's own computer after the model and dependencies have been installed.

That offers several advantages:

More control over data: Product photos and prompts can be processed locally instead of being sent to a hosted AI inference API by the application.
Offline capability: Once installation is complete, the core generation workflow can operate without an internet connection, provided the local services are running.
Less dependence on hosted APIs: The application does not require a paid cloud AI API for its core generation workflow.
Room to experiment: The model and prompting approach can be modified as the application evolves.

Local AI does have trade-offs. Performance depends on the user's hardware, and the model can still make mistakes. Local inference doesn't automatically guarantee that every form of data is erased or that every generated claim is correct.

For me, open innovation made it possible to experiment with an image-capable model as part of a practical application without building the entire product around a hosted AI service.

Prize Categories

Best Use of Gemma : ItemPilot uses Gemma 3 4B as its core vision-language model for analyzing product photos and generating listing content locally through Ollama.

Wingscape: Bringing Urban Birds Closer with Open-Weight AI

Sanjay Kumar Mahto — Sun, 11 Oct 2026 14:10:38 +0000

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

What I Built

What if an AI-powered app could inspire you to spend less time on your screen and more time in nature? 🌿

That question inspired me to build Wingscape — Discover. Observe. Welcome.

Living in cities and apartments doesn't mean we have to lose our connection with nature. Birds visit balconies, neighborhood gardens, rooftops, and nearby trees every day, but we often overlook them.

Wingscape brings together bird discovery, an AI-powered balcony coach, and a simple bird-sighting journal to help people reconnect with the wildlife around them.

The idea is simple: use technology to take the first step, then put the screen aside.

Because sometimes, the best thing an app can do is give you a reason to go outside. 🐦

🐦 Discover Birds
The bird guide introduces eight common Indian urban bird species, including the House Sparrow, Common Myna, Rock Pigeon, and Rose-ringed Parakeet.

Users can search by common or scientific name, explore identifying characteristics and habitats, view photographs with attribution, and learn responsible observation tips.

🌿 AI Balcony Coach
Not everyone has a garden. Sometimes all you have is a small balcony, a few pots, and the desire to make that space more welcoming to birds.

The AI Balcony Coach takes the user's balcony size, sunlight, existing plants, goals, and budget into account, then generates practical recommendations tailored to those preferences.

📔 My Sightings
I also built a simple bird-observation journal where users can record a species, observation date, number of birds, and optional notes.

They can review their sightings and download the records as a CSV file.

NOTE: The current journal uses temporary session storage, so downloading the CSV is important if someone wants to keep a copy of their observations.

Demo

🌿 Live application: https://wingscape.streamlit.app

The app is deployed on Streamlit Community Cloud.

You can explore the bird guide, generate an AI-powered balcony plan, and try the sightings journal directly in your browser.

Code

sanjay067-cse / Wingscape

Discover, observe, and welcome birds to urban spaces.

🌿 Wingscape — Discover. Observe. Welcome.

Bring nature closer to home.

Wingscape is a nature-inspired web application that encourages people living in cities and apartments to discover local birds, make their balconies more bird-friendly, and build a habit of observing the natural world around them.

🔗 Live Demo: https://wingscape.streamlit.app
💻 GitHub Repository: sanjay067-cse/Wingscape

🌱 About the Project

Living in a city doesn't mean being disconnected from nature. Birds visit balconies, neighborhood gardens, rooftops, and nearby trees every day, but we often overlook them.

Wingscape aims to make these small encounters easier to notice and more meaningful.

It combines a curated guide to common Indian urban birds, an AI-powered balcony planning assistant, and a personal bird-sighting journal in one simple, interactive application.

The goal is simple: help people discover the wildlife already around them and take small, responsible steps to welcome it.

✨ Features

🐦 1. Discover Birds

Explore eight common…

View on GitHub

The repository contains the Streamlit application, the bird species dataset, the AI Balcony Coach, the sightings journal, and the dependency configuration.

How I Built It

I built Wingscape using Python and Streamlit, with a calm, cream-and-green design inspired by a nature journal.

The bird information is stored in a JSON dataset, while a separate helper module handles bird-detail rendering. The My Sightings feature lets users record their observations and export them as a CSV file.

For the AI Balcony Coach, I integrated the GPT-OSS 20B open-weight model through the Groq API. The application sends the user's balcony preferences to the model and displays personalized recommendations for making their balcony more welcoming to birds.

The model uses hosted inference rather than running locally on my laptop. I chose this approach so I could build and test a working AI feature without needing the hardware resources required to run a large model locally.

Architecture

The diagram below illustrates how Wingscape's main components work together, from deployment and bird discovery to AI-powered recommendations and sightings export.

The application is deployed through Streamlit Community Cloud using the code hosted on GitHub. The AI Balcony Coach accesses the Groq API using a secret API key, while the bird-discovery and sightings features handle bird information and user observations. Sightings are stored temporarily in the session rather than in a permanent database.

I built Wingscape around a simple idea:
Technology can be more than a screen we stare at. It can be a gentle reminder to step outside, reconnect with nature, and discover the little wonders living just beyond our windows. 🌿

By turning personalized balcony recommendations into small, practical actions, I hope to help people notice, welcome, and appreciate the birds sharing their urban spaces.

Why Does Open Innovation Matter?

Open innovation matters because powerful technology should create opportunities for more people to build, experiment, and solve problems—not just use what someone else has already built.

For Wingscape, I chose the GPT-OSS 20B open-weight model because I wanted to explore what I could build with a model whose weights are openly available, rather than depending exclusively on a proprietary model with inaccessible weights.

A closed API can also generate balcony recommendations. What matters to me is the flexibility open-weight models offer: the ability to explore different inference providers, investigate alternative deployment options, and potentially run or adapt a model independently, subject to its licence, hardware requirements, and other practical constraints.

Wingscape currently uses Groq's hosted inference, so its AI-powered recommendations require an internet connection. I haven't fine-tuned the model or implemented offline inference. But starting with an open-weight model gives me room to learn, experiment, and explore those possibilities as the project evolves.

More importantly, open innovation gives individual developers like me a chance to turn existing advances in AI into something meaningful. I didn't need to build a model from scratch to explore a practical problem: helping people living in urban spaces make their balconies more welcoming to birds.

And that brings me to the idea behind Wingscape: technology should be a bridge to the world beyond our screens, not a reason to stay glued to them. 🌿

I want AI to do more than generate answers. I want it to inspire curiosity, encourage small real-world actions, and help people reconnect with the nature already around them.

My Agent Session

I used Antigravity as a coding assistant while building and debugging Wingscape, working incrementally to implement the bird guide, navigation, AI integration, and sightings journal.

Closing Thoughts

Wingscape started with a simple question:
What if technology helped us pay more attention to nature instead of keeping us in front of a screen?

This is an early version, and there is still plenty to improve. For now, I'm happy to have a working application that combines bird discovery, AI-powered balcony suggestions, and a basic observation journal.

My next goal is to keep improving the experience while encouraging more people to notice and welcome the birds around them.

Discover. Observe. Welcome. 🌿