Inspiration
As the tech industry continues to grow, the number of candidates applying for any given job is increasing exponentially. With so many applicants to choose from, it can be difficult for employers to accurately assess each candidate's qualifications and make informed hiring decisions. In the past, interviewers relied solely on a candidate's technical skills to determine their suitability for a job. However, in today's competitive job market, technical skills are no longer enough to set a candidate apart from the rest. At uPrep, we understand the importance of accurate personality metrics in the tech industry. Our AI-powered platform uses co:here to assess a candidate's soft skills and provide employers with detailed insights into their personality and potential for success.
What it does
Our team utilized co:here's API to leverage AI technology and assess candidates' technical communication skills, speech flow, and professionalism. To access this tool, users are required to create an account on our platform and start solving posted problems. When the problem prompt is opened, a speech-to-text interface is activated with the question prompt displayed on the screen. The user then answers the question with their voice within a 3-minute time frame. Once the interview is concluded, co:here's API calculates an overall grade for each question and provides constructive feedback to the user.
How we built it
We developed a bespoke classification model using curated data sets to analyze voice explanations for existing Leetcode questions. Leveraging the co:here API, our model evaluates user responses and compares them to pre-defined "good" and "bad" responses in the data set, generating an overall grade for the user's performance in the mock interview. In addition to the overall grade, the API provides a detailed breakdown of any mistakes made during the interview, enabling users to gain valuable insights into their communication style and refine their technical communication skills.
Challenges we ran into
During the course of our project, we faced a significant challenge in the development of our classification model. Our initial use of a pre-built model presented challenges in setting appropriate parameters and generating realistic mock interviews. To overcome this obstacle, we built a custom classification model from the ground up, tailored to our specific parameters and training data. As the project deadline was tight, we had to quickly gain a comprehensive understanding of co:here's API in under 40 hours. Ultimately, our team's dedication and expertise allowed us to develop a robust and effective solution, delivering high-quality results for our demo.
Accomplishments that we're proud of
The uPrep team has developed a robust data set generation system that efficiently stores data within the MongoDB framework. We take pride in our team's innovation and our ability to create metrics that quantify previously "intangible" skills.
What we learned
As a team, we gained valuable experience and expertise in a range of technologies during the development of our product, including Django, React, and AI. Additionally, we developed skills in product design and the process of swapping technologies. These experiences allowed us to broaden our capabilities and knowledge, enabling us to deliver high-quality solutions for our clients. We look forward to leveraging these skills in future projects and continuing our growth and development as a team.
What's next
Our goal is to develop a highly realistic model that brings uPrep even closer to providing life-like interviews, in addition to our current feedback capabilities. Additionally, we aim to enhance uPrep by introducing premium features and an "enterprise" version for businesses to use in their own interview screening processes. To ensure the wide adoption of our platform, we will work to increase the marketability of uPrep scores to recruiters in order to integrate them in resumes and application processes. Through these efforts, we will continue to innovate and deliver value to our clients, driving the success of uPrep and solidifying our position as a leader in the industry.
Log in or sign up for Devpost to join the conversation.