Spark

Ignite your potential! Design, Practice, Analyze your interviews

Comment

Inspiration

Interview season has been coming around for many months now and we were inspired to create a web application that is able to help students, especially ones in college like Purdue University, to practice their elevator pitches and interviews with a virtual assistant at the touch of their fingertips.

What it does

Spark is an AI-powered application that provides an interface for users to input their desired career/job descriptions, resume, and other details to generate a personalized interview for them. Users are able to then have a near-real-life conversational interview with a virtual assistant. The application will also analyze interview results and speech to detect various patterns and areas of improvements for the interviewees to work on.

How we built it

Spark was built using NextJS for the front-end and NodeJS + Python for the back-end server and processing respectively. Our speech analysis models were built with TensorFlow.

Challenges we ran into

One of the largest challenges we've had while working on the project was creating and modifying the front-end UI sufficiently such that it was a pleasant and easy to use experience for all. Connecting this front-end to the back-end required an extensive amount of effort and work to be done, with numerous hours of testing to ensure that all systems were functioning as they were designed and there are minimal bugs and issues to be encountered.

Accomplishments that we're proud of

We are proud of everything we have accomplished with this project. Compared to our previous 2 years' projects, this one was a major undertaking and far more extensive in both breadth and content by far. We have had to use many new frameworks, algorithms, and methods to achieve our final result and we are proud to have completed it on time.

What we learned

We have learned to use NextJS to build UI frameworks, TensorFlow to train speech emotional analysis models, and philosophies of front-and-back-end design ensuring that all parts of the process are able to communicate sufficiently and efficiently.

What's next for Spark

We hope to continue developing Spark, adding improvements (video chat!) and more analysis features for users. Potentially, this project could be implemented further and made commercially viable.

Built With

Share this project:

Updates