DukaanAI
An AI-assisted inventory management tool aimed at small retailers and local shop owners. It automatically generates product names, categories and descriptions from uploaded product photographs, with an emphasis on local-first processing and human verification.
Timeline (last 90 days)
DukaanAI is an AI-assisted inventory tool for small retailers and local shopkeepers, whose core function is that after uploading a product image, the AI generates a suggested product name, category, and description, and the shopkeeper confirms and adds the selling price and stock before saving it to inventory.
DukaanAI's MVP includes six features: product image upload, AI-generated name/category/description, editable fields, manual entry of selling price and stock, inventory search and deletion, and CSV catalog export.
DukaanAI's frontend is built with React and Vite, and its backend uses Node.js and Express.
DukaanAI's core AI capability comes from the locally running Gemma 3 4B model, invoked through Ollama, which analyzes the image and returns structured JSON.
DukaanAI's initial version stores inventory data in the browser's localStorage, making it a local-first prototype rather than a cloud-based multi-user system.
DukaanAI's core trade-offs in choosing an open-weight model and local inference are: the marginal cost of inference is close to zero and product data does not leave the local device, but the shopkeeper needs suitable hardware, inference speed may be slower, and the model output may contain errors.
DukaanAI's product design principle is that AI generates suggestions and humans confirm, rather than fully automatically replacing the shopkeeper's judgment, to compensate for the limitations of local inference.
DukaanAI's initial target users are independent grocery stores and small retailers that need a simple way to create product catalogs.
The developer plans to verify through testing with local shopkeepers whether DukaanAI truly saves time, and may adopt a monthly subscription business model, but willingness to pay needs to be verified before pricing.
DukaanAI's future improvement directions include cloud sync, store accounts, barcode support, low-stock reminders, and integration with existing point-of-sale systems.
1 more timeline events
All Facts (10)
DukaanAI's inventory data is initially stored in the browser's localStorage, and once multi-device sync or multi-person collaboration is needed, it must be migrated to a cloud database.
50%UnverifiedDukaanAI's MVP includes six features: product image upload, AI-generated name/category/description, editable fields, manual entry of selling price and stock, inventory search and deletion, and CSV catalog export.
50%UnverifiedDukaanAI's frontend is built with React and Vite, and its backend uses Node.js and Express.
50%UnverifiedDukaanAI's core AI capability comes from the locally running Gemma 3 4B model, invoked through Ollama, which analyzes the image and returns structured JSON.
50%UnverifiedDukaanAI's initial version stores inventory data in the browser's localStorage, making it a local-first prototype rather than a cloud-based multi-user system.
50%UnverifiedDukaanAI's core trade-offs in choosing an open-weight model and local inference are: the marginal cost of inference is close to zero and product data does not leave the local device, but the shopkeeper needs suitable hardware, inference speed may be slower, and the model output may contain errors.
50%UnverifiedDukaanAI's product design principle is that AI generates suggestions and humans confirm, rather than fully automatically replacing the shopkeeper's judgment, to compensate for the limitations of local inference.
50%UnverifiedDukaanAI's initial target users are independent grocery stores and small retailers that need a simple way to create product catalogs.
50%UnverifiedDukaanAI is an AI-assisted inventory tool for small retailers and local shopkeepers, whose core function is that after uploading a product image, the AI generates a suggested product name, category, and description, and the shopkeeper confirms and adds the selling price and stock before saving it to inventory.
50%UnverifiedDukaanAI's future improvement directions include cloud sync, store accounts, barcode support, low-stock reminders, and integration with existing point-of-sale systems.
50%