ToyoTrack

Creating user profile/authentication
explore/browse page
filters/preferences
payment calculator/financing
payment calculator/financing cont.
compare two vehicles
leasing payment calculator

Inspiration

We were inspired to create ToyoTrack we love Toyota's reliability and car shopping and financing less intimidating. Going to a dealership right off the bat can be overwhelming with big terms, pushy salespeople, and complex loan terms so we wanted a tool that explains financing options in an easily understandable way. With our AI chat box, people can ask any question they want about cars or payments without pressure or judgment, making the whole experience more enjoyable.

What it does

ToyoTrack lets users explore vehicles and financing options by easily filtering to match what you're looking for on your budget. Our AI chat box allows for flexible research and accessible answers. You can ask questions like “Compare TRD Off-Road vs TRD Pro”, “Find me a SUV within a 40k budget”, or "What financing plans can I get for a corolla with a poor credit?" and the app maps that to structured vehicle data. It pulls from a curated JSON dataset of cars and trims, runs queries on the backend, and returns clear comparisons, summaries, and key specs. The goal is to make browsing options and understanding financing more conversational and efficient.

How we built it

We designed reusable and user friendly UI components using figma and with react, typescript, and HTML, we brought to life a clean and easy to navigate front end. Using node.js and express, we carefully programmed APIs and logic that connect our front end components to their actual functionalities, while also handling user/session authentication. For the back end portion, we used PostgreSQL to create scalable data tables to store user models and car specifications taken from CarsXE API. We also incorporated a chat box programmed using the Gemini AI API.

Challenges we ran into

One big challenge we had was working with the Gemini API while keeping its answers tied to our own car and financing data. We had to carefully design system prompts, give clear user instructions, and treat the AI as stateless while still passing in enough conversation context each time. Another challenge was the car data layer itself: we had to use Claude to help parse large JSON files into cleaner, usable data structures.

Accomplishments that we're proud of

We’re really proud of how natural and interactive the chat box feels, making it easy for users to explore cars and financing options. The interface is user-friendly, and the payment logic is presented in a way that’s simple to follow. We’re also happy with our authentication and session handling, which reliably stores user data and keeps the experience personalized and secure.

What we learned

We learned how to integrate existing APIs as well as design our own APIs to implement custom logic on the backend. We also gained hands-on experience with user authentication and session management. We also thought about how to make the product both functional and marketable, especially in how we present leasing and financing options clearly for users.

What's next for ToyoTrack

Some ideas we have for the future include expanding to the sales of both used and new cars as well at training an AI finance-focused consultant who can work with users to figure out more personalized planning.