Describe what you want to make
Customers explain their DIY plans in their own words. The assistant asks relevant questions about the project, dimensions and finish.
BricoGuide
From a DIY idea to a practical materials list
2026
BricoGuide is a proof of concept for DIY customers who know what they want to make but are not sure where to start. Through a conversation, the app asks about the project, suggests materials and quantities, and gives an estimated total based on demo product data. Customers can also upload photos for additional context and ask a store employee to review the proposal. A separate dashboard connects the employee to the customer's project in real time.
- role
- Concept, UX/UI Design, Frontend & Backend Development, AI Integration
- type
- AI-Assisted DIY Retail Prototype
From idea to first proposal
Photos for extra context
An inspiration image or photo of the space can provide context. When AI is available, image analysis helps clarify the project.
Materials and cost estimate
The app prepares an initial list of products, quantities and an estimated total. The prototype uses demo data, not a live retail catalogue.
A store employee can step in
Customers can request advice. A store employee sees the request in a separate dashboard and can respond and review the materials list.
Proposal to demo cart
The proposed materials can be collected in a demo cart. It illustrates the next shopping step without placing an actual order.
Why I combined AI with human advice
I did not want a chatbot that simply spits out a list of products. The assistant first helps customers clarify what they need. But DIY projects can involve missing details and safety-related decisions. That is why a store employee can review and adjust the proposal. AI helps start the conversation; a person remains available when it matters.
One project, two connected interfaces
The customer interface focuses on a simple conversation and a clear project overview. The employee dashboard shows open requests, project details and chat history. Real-time messages connect both sides. Behind the scenes, the backend, session storage, structured question flows and Gemini work together to build a proposal from the customer's input.
technologies
Frontend
- Next.js
- React
- TypeScript
- Tailwind CSS
Backend and real-time
- Node.js
- Express.js
- Socket.IO
AI and data
- Google Gemini
- JSON demo catalogue
- Image analysis