Explain This Thing: An AI That Helps My Parents Understand Their Devices
This is a submission for the Hacktoberfest Weekend Challenge: Build for a Friend What I Built My parents often use household devices like the AC, microwave, TV, and dishwasher, but they sometimes forget what particular buttons do. Usually, they would have to ask me: “What does this button do?” And because these devices have so many buttons and modes, remembering them isn't always easy. So I built…
This submission for the Hacktoberfest Weekend Challenge is a project called "Explain This Thing" designed to help a user's parents understand the functions of household devices such as the AC, dishwasher, microwave, and TV remote. The AI assistant allows users to upload photos of a remote or control panel or use a saved photo, then tap the button they are confused about to receive a simple explanation in English or Hindi.
The goal was not to build an AI that understands every appliance but to focus on the few devices the parents actually use. The app currently supports four devices: Bedroom AC remote, dishwasher, Panasonic microwave, and Samsung TV remote. After giving the app to his parents and asking them to try it, the user found that it helped them understand the buttons without needing external help.
They also appreciated being able to quickly look up the meaning of certain buttons themselves, which solved the main problem the project aimed to address. The project is built using React and Vite for the frontend, and Python FastAPI and RapidOCR for the backend. The AI runs locally using Ollama and Gemma 3 4B for vision and explanation, as well as nomic-embed-text for embeddings.
The code is available on GitHub. The project emphasizes the importance of open innovation and using an open-weight model running locally instead of relying on a closed AI API. This provides more control over data processing, eliminates the need for API requests, and allows for experimentation and customization.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
