Toyo

Inspiration

Buying a car can feel overwhelming: comparing models, understanding financing, and figuring out what you can comfortably afford. We wanted to make that process more approachable. Toyo turns car shopping into a guided journey through a playful 3D town, with an AI companion that helps users explore their options one step at a time.

What it does

Toyo is an interactive Toyota shopping assistant. Users explore a pink 3D town, discover vehicles that match their preferences, and build an illustrative ownership plan.

Through voice or on-screen controls, users can:

• Compare catalog vehicles by category and budget.
• Explore financing while accounting for emergency savings, running costs, and insurance.
• Search for current prices and offers through linked web results.
• Complete a simulated checkout and save a demo reservation.
• Navigate supported steps using an Arduino button.

A separate staff dashboard supports reservation review, status updates, and exports. Purchases, financing, insurance, and reservations remain demonstrations rather than real transactions.

How we built it

We built the interface with React and TypeScript, using Three.js and React Three Fiber for the town and robot companion. A Node.js and Express backend manages the shopping journey, validates actions, and calculates financing.

Gemini supports vehicle recommendations and real-time voice conversations. SerpAPI provides price and offer research links. Reservation storage supports local SQLite or Tiger Data PostgreSQL, while Arduino integration adds a physical control through browser USB or a local serial bridge.

Financial calculations run in backend code, keeping prices and affordability results consistent across voice, browser, and hardware interactions.

Challenges we ran into

One of the biggest challenges was keeping voice conversations, browser controls, and Arduino input synchronized. A delayed response or button press should never act on an outdated vehicle selection or financial plan.

We also needed to make conversational input precise enough for financial details and contact information. We added validation, readbacks, and confirmation steps so users can review important details before continuing.

Another challenge was clearly distinguishing catalog prices, illustrative ownership estimates, and external search results without making the experience difficult to follow.

Accomplishments that we're proud of

We brought a 3D interface, conversational AI, physical controls, and a reservation workflow into one connected experience.

We’re especially proud that Toyo considers more than a monthly loan payment. Emergency savings, running costs, and insurance help users explore whether a plan fits their broader budget. When it does not, the journey can finish with a savings plan.

What we learned

We learned that a useful AI assistant needs dependable application logic behind the conversation. Voice makes the experience approachable, while validation and explicit confirmations keep actions predictable.

We also learned how much coordination a multimodal interface requires. Every input method needs to share the same current state, and users need clear feedback about what happened and what comes next.

What's next for Toyo

We want to connect Toyo to verified dealer inventory and localized pricing, improve accessibility, and test voice and Arduino interactions with more users and physical hardware.

We also plan to strengthen reservation persistence and staff account management, then evaluate dealership integrations that could turn a reviewed demo plan into a real follow-up request with the buyer’s permission.

Built With

• arduino
• c++
• gemini
• javascript
• react
• timescaledb