🛍️ DukaanAI: Building an AI Marketing Assistant for Local Shops with Open-Weight AI

What I Built 🛍️ DukaanAI — AI Marketing Assistant for Local Shops Small shop owners often struggle to create attractive advertisements for their products. They may not have the time, marketing skills, or budget to hire someone to write promotional content. DukaanAI is a simple AI-powered project that helps local shop owners generate product advertisements quickly. For example, a shop owner can ask it to create an advertisement for a cotton kurti priced at ₹599 in Marathi. The goal is to make AI-powered marketing more accessible to small Indian businesses. Key Features Generate product advertisements from a simple text prompt. Create promotional content for different products and prices. Support multilingual marketing prompts, including English, Hindi, and Marathi. Help small businesses create content for WhatsApp and social media. Keep the project simple and easy for beginner developers to understand and extend. The initial version focuses on generating advertisements. More features can be added as the project develops. Who would pay for it? Local shop owners, home-based businesses, small retailers, and online sellers could benefit from a tool like DukaanAI. A future business model could offer free basic generations and paid bulk advertisement creation. Demo The project currently runs as a simple Python application. Demo video: Watch Demo Video Live demo: Not deployed yet. Code GitHub repository: [Add your GitHub repository link here] The project uses Python and the Groq API to generate marketing content with an open-weight AI model. How I Built It I built DukaanAI using Python and the Groq API. The basic workflow is: The user enters a product and a request for an advertisement. Python sends the prompt to the AI model through the API. The model generates promotional content. The application displays the generated advertisement in the terminal. Technology Stack Python: Application logic. Groq API: Access to hosted model inference. Open-weight model: openai/gpt-oss-20b. Groq Python SDK: Connects Python to the model API. I chose a simple implementation so that I could focus on solving a real problem rather than building unnecessary complexity. Why Does Open Innovation Matter? Open innovation makes powerful AI models more accessible to independent developers and small businesses. For DukaanAI, using an open-weight model means the underlying model weights are available under their applicable license, allowing developers to explore customization and alternative inference arrangements. Using hosted inference through Groq also makes it possible to experiment without downloading or running a model locally. This approach has trade-offs: API access may have usage limits or costs, and product information sent to a hosted service is processed by that provider. In the future, DukaanAI could support local inference or other compatible providers, giving users more flexibility and reducing dependence on a single service. I believe open AI technology can help developers build affordable, practical tools for people who might otherwise have limited access to AI. My First User This project is still in its initial stage. I have not yet confirmed a real-world user or received payment. My next goal is to let a local shop owner try DukaanAI, collect feedback, and improve the generated advertisements based on their needs. Prize Categories Potential category: Best Use of a qualifying open-weight AI model. I will confirm the applicable partner category against the official challenge rules before entering it. Thanks to the Hacktoberfest and DEV communities for encouraging developers to build useful projects with open AI technology!