The basic motivation for this project was to create a tool that would simplify the interview process and automate candidate selection for a job by removing human biases through a more objective framework AI and Machine Learning.
What it does
ough the project is in its very initial phase. The basic premise of this project is to take a prerecorded interview video or audio and upload it to Assembly AI through which it will return a text transcript. Then it will be further analyzed through other NLP AIs which will evaluate the candidates skills, experiences, "ranking" in comparison to other candidates (also based on custom criteria given by the company) and finally yield a qualitative and visual answer on a candidates employability (ie. whether someone is a good fit for the company or not) Please see the chart below for a better visualization of the process.
How we built it
The technology used creating this project is both simple and complex. First the transcription was done through Assembly AI. The webapp was created using Streamlit (Python tool). The text analysis was done through Parallel dots API.
Challenges we ran into
I ran into some challenges with the Assembly AI. However, the documentation helped me resolved the issue.
Accomplishments that we're proud ofHaving a working prototype to showcase for hackathon.
What we learned
Working with this project I learned how to work with Assembly AI. I also learned how to use Streamlit to make quick web APPs. Additionally, I gained a deeper understanding of NLP and AI
What's next for uhub
Adding more complex machine learning, NLP and Custom Algorithm for better text analysis Improving the UI for professional use. Adding youtube_dl functionality so video can be directly inputted through YT And many more