-
-
TOYOTA Initial Diagnostics
-
Inspiration
As a Toyota owner, I personally experienced how difficult it can be to find an affordable vehicle that meets specific needs like my budget, fuel efficiency/gas costs, maintenance costs, and availability which often don’t line up. Many car marketplaces fail when inventory is limited or filters are too strict. So, I developed this hack Toyota Initial Diagnostics to help those customers in need of a new car by providing meaningful results with every query.
What it does
This e-commerce web application helps you filter all cars by Price, Make, Model, etc. It also recommends cars upon zero-match that are similar to your chosen filters in the order of active filters. As well as a handy chatbot Initial Assistant that can apply filters and clear them on your behalf. It is an initial diagnostic to your new ride!
How I built it
Next, React.js web app with Gemini APIs, recommender system, & diagnostics scripts.
Challenges I ran into
A lot of Gemini API errors, but luckily the API documentation was updated and I was able to configure it. I had more trouble with the recommender logic.
Accomplishments that I am proud of
I was able to make the recommender be more dynamic taking in substitute data and my mock data to make real time recommendations.
What I learned
I learned more about recommender systems, and while mine doesn't have a custom ML engine, it does use an LLM to help learn the users needs. I hope to continue learning more about these systems and there long term use and scalability
What's next for Toyota Initial Diagnostics
- Improve recommender system with more data.
- Use APIs to fetch live car data.
- Add MongoDB or another database to store the data, as well as an admin dash to monitor sales and add/edit/remove products.
- Add accounts for users to save their preferences, this will help the system learn over time.
- Add checkout system with bank wire, credit, etc. for product purchase.
- Deploy to Vercel! or Cloud!
Log in or sign up for Devpost to join the conversation.