DriveWise Project

Inspiration

The idea for DriveWise came from my own personal experience of buying a car after graduating high school. At the time, I had managed to save enough money to purchase my first car, and I anticipated the process would be simple and straightforward. However, I quickly realized that the car-buying journey was far more complicated than I expected, with numerous variables to consider, such as budget constraints, financing options, car features, and long-term commitments. I found myself overwhelmed by the choices and the complexity of the decision-making process.

This experience sparked the idea to create a tool that could help others navigate this daunting process with ease. I wanted to build an AI-powered system that could take personal details into account and recommend the best car for each individual based on their unique financial situation and preferences. Thus, DriveWise was born—a full-stack AI agent-based application designed to help people find the best car suited to their needs.

What It Does

The DriveWise application collects detailed user input to provide personalized car recommendations. Key factors such as:

• Age
• Marital status
• Family dynamics
• Income
• Credit score
• Monthly payment budget
• Down payment budget

are gathered and processed to match users with the best car options based on their financial situation and personal needs. The system uses these inputs to suggest cars that are affordable within the user’s budget, as well as those that fit their lifestyle requirements (e.g., family-friendly cars, fuel-efficient options, etc.).

How We Built It

We chose a full-stack approach for building DriveWise to provide a seamless user experience. The backend is powered by Flask and Python, allowing for flexibility and robust handling of the AI logic and data management. For the frontend, we used React to create an interactive and responsive interface, while Tailwind and Material UI (MUI) were employed to ensure a clean, user-friendly design.

Firebase was integrated to handle user authentication and store chat histories, ensuring secure and reliable data management. The real-time chat system allows users to engage with the application dynamically, getting instant feedback and support while exploring their car options.

The AI engine is designed to handle large amounts of user data and run sophisticated recommendation algorithms to ensure the best matches. We leveraged Phidata for AI agent management after our initial struggles with CrewAI.

Challenges We Ran Into

One of the major challenges we faced was developing and training the AI agents. We initially chose CrewAI for the AI agent but encountered issues with integrating it effectively into our workflow. This led to a pivotal moment where we had to pivot and explore alternative frameworks. After evaluating several options, we decided to move to Phidata, which provided the flexibility and scalability we needed for managing complex data and improving recommendation accuracy.

Another challenge was ensuring that the car recommendations were not just financially viable, but also aligned with users’ specific preferences and lifestyle. Balancing all these factors while keeping the application intuitive was a continuous learning process.

Accomplishments We're Proud Of

Despite the challenges, we're incredibly proud of what we've built. We managed to create a customer-first application that takes a holistic approach to the car-buying process. By focusing on both the financial and lifestyle needs of users, we were able to design an intelligent recommendation system that truly adds value.

Additionally, our ability to integrate AI agents into the decision-making process was a major achievement. DriveWise offers users a personalized experience that guides them through a complex process with ease. The real-time chat system also provides a human touch, further enhancing the user experience.

What We Learned

Throughout this project, we learned the importance of adaptability in AI development. The challenges with agent frameworks underscored the need to remain flexible when building complex systems. We also gained deep insights into how AI recommendation engines work, and how to balance multiple factors (financial, personal, and social) to generate meaningful recommendations.

Additionally, I refined my front-end development skills, particularly in working with React and Tailwind. Building an intuitive interface that’s easy for users to navigate was a crucial learning experience that will continue to influence future updates to the platform.

What's Next for DriveWise

Moving forward, we have a lot of exciting plans for DriveWise. We want to enhance the car comparison feature by offering more detailed insights into vehicle performance, safety ratings, and long-term cost projections. Additionally, we aim to incorporate a price fairness engine that will compare prices of similar cars in the market to ensure users get the best deals.

Another key area of development is advanced model comparison. Right now, users can compare basic features between two cars, but we plan to refine this to include more granular comparisons—such as the overall cost of ownership, financing terms, and how different models impact long-term financial goals.

Built With

• openai
• python
• react