This repository is arranged as 3 submodules consisting on Vision, NLP and rule based models. The folder structure of the repo is as follows:
project_root/
├── data/
│ ├── raw/
│ ├── processed/
│ └── external/
├── notebooks/EDA
├── src/
│ ├── preprocessing/
| | ├── data_cleaning/
| | |── data_transformation/
| | |── feature_engineering/
| | └── data_normalization/
│ ├── modeling/
| | |── model_selection/
| | |── model_training/
| | └── model_tuning/
│ └── evaluation/
├── includes/
| |── utilities/
| └── constants/
|
├── models/
| |── vision/
| |── audio/
| └── text/
├── reports/
├── requirements.txt
├── config.yaml
└── README.md
Growing organizations conduct multiple recruitment drives for finding the best candidates for being part of their company. Post pandemic there has been a rise of virtual interviews. During the entire recuitmnent process some of the evaluations and actions taken are rule based eg. experience, expected CTC, notice peroid etc, which can be automated. There are some criteria which are not in the books but can cause a candidate to be rejected or selected for next round eg.Language, fluency, tone, etc. A machine can be trained to analyse such traits and provide a candidate's behavioural insights to the interviewer. While other evaluations require human in the loop like scenario based questions, use case implementation etc.
In real time the interview panel used to listen to the recording of initial screening round, before selecting the candidate for next round. There can be over 50 recordings that the manager might have to go through in a day to make this decesion. This is a tidious task. Here in when the AI recruiter could pitch in. It can perform the above tasks and help the management to make smart choices.
This AI solution can provide visual,linguistic and rule based insights from video recordings of virtual meetings.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent