ScreenSage

Sample Question
Main Page

Inspiration

The motivation behind ScreenSage came from the desire to find high-quality and accessible interview preparation tools. We realized that interviews always caused a large amount of anxiety and that there were little to no practical tools that could provide interview feedback as well as assist small businesses with hiring processes. Studies have shown that AI models can be trained to reason in a less biased way. In turn, this is a potential way to eliminate discriminatory biases during interview-like processes. Thus, interviewees can improve their responses for future applications (building confidence!), and the platform would be accessible for any company to use, helping smaller businesses within their hiring processes.

What it does

Screen Sage is a software designed to help both interviewers and interviewees with pre-screening interviews. The software takes a position/role a person is applying to and generates questions to be during a pre-screening interview for the stated position. While running the program, interview questions are generated, and the user is prompted to type out their responses in the software. After the user answers all five questions, the software provides insightful and detailed feedback for each of the user's responses, along with a 'score' out of 100.

Tech stack and languages

ScreenSage was built using a simple static HTML, and CSS front-end. To implement Cohere's API, we utilized Flask.

Challenges we ran into

As beginners, we were unfamiliar with using Flask as a front-end framework to integrate Cohere easily. We initially thought of using React to integrate a back-end. However, this turned out to be daunting, and we ended up reverting to Flask.

Accomplishments that we're proud of

This turned out to be a great learning opportunity for all of us! As we were all unfamiliar with Flask, this provided a great opportunity to learn how to integrate APIs and HTML methods such as GET and POST. We're also glad that the front end is simple, nice, minimalistic, and accessible for users.

What we learned

We definitely learnt a lot during this project! More especially, about connecting the front end with the back end, as most of us were more comfortable with the front end instead.

What's next for ScreenSage?

What's next? Well, we plan to use Cohere's model in a better fashion to improve the feedback the user receives at the end of the quiz. In fact, this could be in the form of feedback for each question, and the feedback can be tailored based on certain interviewer's preferences.

As for the UI/UX components, making it responsive would be key, as well as giving it a bit more of a 'pop'!