AIvantage | Devpost

Inspiration

We realized how stressful the interview process can be, both from our own experiences and research. In fact, 93% of candidates experience anxiety before interviews, and 41% worry about not being able to answer difficult questions. On the flip side, 96% of candidates who had mock interviews landed their "dream job." That’s why we created AIvantage—a platform to boost confidence and help users refine their interview strategies with personalized training and real-time practice.

What it does

AIvantage is an AI-powered interview prep platform designed to give users a personalized experience. It uses computer vision to track your focus during mock interviews and generates unique real-time metrics that help identify distractions. The platform creates tailored interview questions, adapting to each user’s needs and challenges. Additionally, we’ve implemented a job-scraping feature using Python, allowing users to browse real-time job postings directly within the platform and apply for them with a single click. Users can also transcribe and control the platform through voice commands, enabling hands-free navigation during practice sessions. This all-in-one approach helps boost confidence and prepare users to handle even the toughest interview questions, while seamlessly connecting them with job opportunities.

How we built it

AIvantage was built using React.js for the front end, styled with Tailwind CSS to ensure a clean, responsive interface. We used the Mediapipe library for facial tracking, enabling real-time monitoring of user focus during practice sessions. The Gemini API powers the generation of personalized interview questions, creating a tailored training experience based on user performance. For the job-scraping feature, we used Python to gather live job postings from various websites, integrating them into the platform so users can easily browse and apply for relevant roles. Hosting the platform entirely client-side ensures faster load times and a smooth user experience, while Python runs on the backend to keep the job listings updated.

Challenges we ran into

The biggest challenge was developing a feature-rich application where each user session feels unique. We had to ensure that the personalized questions, focus tracking, and real-time metrics were dynamically adjusted with every use. Additionally, building a reliable scoring system that evaluates user performance based on multiple factors—like attentiveness and accuracy—required extensive testing and fine-tuning.

Accomplishments that we're proud of

We’re proud of learning how to effectively integrate Mediapipe for facial tracking and using the Gemini API to generate personalized questions. Adding the Python-based job scraping functionality was another milestone, allowing users to take immediate action on real job opportunities. Maintaining the entire application client-side, except for Python’s backend job scraping, optimized performance and made hosting easier. The fact that we managed to bring all these features together in a hackathon setting is something we’re incredibly proud of. By doing all of this we were also able to learn more about the job market, and what it takes to be a good interviewee.

What we learned

We learned how to effectively work as a team, distribute tasks, and tackle real-world problems. Our biggest technical takeaways include working with Mediapipe for real-time facial tracking and integrating the Gemini API for AI-driven question generation. We also learned how to balance multiple advanced features while maintaining a seamless user experience.

What's next for AIvantage

Next, we plan to provide deeper insights and even more tailored help by training the platform to learn from user submissions. Over time, the platform will build a "learning relationship" with each user, offering more precise feedback and continuous improvement in interview preparation.