Chatbot‑Dog is a lightweight AI‑powered assistant designed to answer questions and give basic advice about dogs. The goal is to help dog owners get quick, accessible information before consulting a veterinarian if needed. This project was built in a team of four during the final Full‑Stack Web & AI project at Le Wagon, where we had only five days to design and ship an application integrating a Large Language Model. We developed Chatbot‑Dog using Ruby on Rails and RubyLLM to generate dynamic, dog‑focused responses.
Chatbot‑Dog est un assistant IA léger conçu pour répondre aux questions et donner des conseils de base autour des chiens. L’objectif est d’aider les propriétaires à obtenir rapidement des informations accessibles avant, si nécessaire, de consulter un vétérinaire.
Ce projet a été réalisé en équipe de quatre dans le cadre du projet final Full‑Stack Web & AI au Wagon. Nous avions seulement cinq jours pour imaginer, concevoir et livrer une application intégrant un Large Language Model. Nous avons développé Chatbot‑Dog avec Ruby on Rails et RubyLLM, afin de générer des réponses dynamiques centrées sur les chiens.
- Ruby on Rails
- JavaScript
- PostgreSQL
- Turbo Stream
- RubyLLM
- Bootstrap
Chatbot‑Dog keeps the experience simple, friendly, and helpful:
- Dog‑Focused Q&A : Ask questions about health, behavior, training, or daily care
- Basic Advice : Get first‑level guidance before consulting a vet
- Conversational Flow : Natural, dog‑themed responses powered by RubyLLM
- Message History : Keep track of previous questions and answers
- Minimal UI : Clean chat interface with real‑time updates
- Fast Interactions : Lightweight backend for quick response generation
We started by setting up a classic Rails structure with models and controllers to handle chat messages. Then we integrated RubyLLM to generate answers based on user questions. The challenge was to create an assistant that feels helpful, friendly, and focused on dog‑related topics — without pretending to replace professional veterinary advice. Turbo Stream handles real‑time updates, making the chat smooth and responsive. The project was built iteratively over five days: message handling → LLM integration → personality tuning → UI polish → deployment. Each step focused on keeping the experience simple, reliable.
Working on Chatbot‑Dog helped me grow in several areas:
- Integrating LLMs into a Rails backend
- Designing conversational flows and assistant “personality”
- Building real‑time chat interfaces with Turbo Stream
- Structuring clean Ruby services for AI communication
- Collaborating efficiently in a fast‑paced team project
- Introduce voice input and audio responses
- Add a knowledge base for more accurate answers
- Improve the UI with animations or a more playful design
- Add user accounts to save long‑term conversations
- Create a mobile‑friendly version with richer interactions
