Inspiration:

GetHired.ai was born from our recent internship season experiences🤞, where mutual support in mock interviews highlighted the need for a tool to aid individual preparation during busy times. Inspired by a prior project at DeltaHacks, we aimed to enhance it with advanced features like semantic and sentiment analysis, as well as video and voice recognition. Our vision is to offer a user-friendly platform that empowers individuals to excel in interviews, even when personal assistance might be limited. GetHired.ai embodies our commitment to making job preparation more accessible and effective for everyone, everywhere!

What it does:

GetHired.ai revolutionizes interview preparation with simplicity and efficiency. Upon login, users select their industry, paste the job description, and receive personalized interview questions. The platform analyzes responses, providing detailed feedback on speech coherence, confidence, body language efficacy, and pause frequency. Notably, feedback is generated in both written and audio formats, reflecting our commitment to reducing bandwidth and ensuring a user-friendly experience. GetHired.ai empowers individuals to effortlessly enhance their interview skills for a successful and confident job application process.

How we built it:

The platform employs a VGG-styled CNN to analyze video frames, evaluate body language, eye contact, and facial expressions, and capture positive emotions like enthusiasm. Speech undergoes assessment through a semantic and sentiment analyzer utilizing a KNN classifier to measure coherence and confidence. Crafted on Next.js, our UI ensures a user-friendly journey. Inspired by Iron Man's Jarvis, responses are generated as voice for an interactive touch. Deployment on Vercel and Azure enhances scalability and performance, making GetHired.ai an innovative and engaging tool for job preparation, education, and productivity🙂!

Challenges we ran into:

Building GetHired.ai came with its fair share of challenges. One major hurdle was trying to train the model on our laptops, which didn't have enough RAM. After three failed attempts and crashes, we finally got a Virtual VM with GPUs after a grueling six hours. Integrating four different machine learning models for simultaneous use in the app was also tricky and required careful coordination. On top of that, we had to brainstorm and implement the entire idea quickly. These challenges taught us a lot about optimizing resources, problem-solving, and working efficiently under pressure and was overall a very valuable experience!

Accomplishments that we're proud of:

We're proud because this was the first time we brought together a bunch of different machine learning stuff and APIs all at once. Video analysis was new to us, and it took some time to figure out, but we're really happy with how it turned out. Plus, for some of our team, Deltahacks was their first hackathon, so they learned a lot. And let's not forget the adventure of commuting from Toronto to Hamilton—it was quite a journey!

What we learned:

The main takeaway: don't try training big models on regular laptops! We learned the hard way. On a brighter note, we discovered that video analysis is a fascinating research field, and we're eager to explore it further in future projects. Learning from our experiences, we'll be more intentional about exposing ourselves to new areas like video analysis to enhance our skills and capabilities in upcoming endeavors.

What's next for GetHired.ai:

Looking ahead, our plans for GetHired.ai involve expanding its utility to other industries, starting with engineering. We're looking to incorporate a technical interview section, featuring questions from platforms like LeetCode and systems design interview queries. Additionally, our focus will be on enhancing the user interface for a more seamless and user-friendly experience. The goal is to continually evolve and refine GetHired.ai to meet the diverse needs of users preparing for a variety of interviews across different domains!

Built With

  • cnns
  • cohere
  • flask
  • keras
  • knn
  • nextjs
  • pytorch
  • semanticanalysis
  • sentimentanalysis
  • speechrecognition
  • sqlite
  • tensorflow
Share this project:

Updates