AutoHRise: AI-Powered Hiring Assistant

AutoHRise An AI-powered recruitment agentic assistant automates the hiring process, reducing time-to-hire and improving candidate experience. The AI agent acts as an intelligent assistant, interacting with recruiters, candidates, and HR systems to streamline hiring.

What is Agentic AI?:

Agentic AI refers to AI that can act independently, make decisions, and complete tasks on its own. AutoHRise is an agentic AI, meaning it doesn’t just provide suggestions—it takes action. It can create job descriptions, post jobs, screen resumes, and schedule interviews without requiring constant human input.

How It Works:

1. Job Posting:

The job posting agent takes input from the user, like required skills and years of experience. Based on this, it creates a job description that can be shared on platforms like LinkedIn and job portals.

2. Resume Screening:

The Resume Screening agent takes input from the user, such as required skills, and finds the top 10 most relevant resumes from Watsonx Discovery. These resumes are then sent to Watsonx.ai, where the "ibm/granite-3-2b-instruct" model extracts key details like the candidate’s name, email ID, mobile number, and skills.

3. Interview Scheduling:

The Interview Scheduling Agent takes input from the Resume Screening Agent and creates a sample email to inform the candidate about the interview schedule.

Tools:

1. watsonx Discovery
2. watsonx AI
3. Framework - Crew AI

Model:

Granite - "ibm/granite-3-2b-instruct"

Getting Started

To get started with AutoHRise, clone this repository to your local machine:

git clone https://github.ibm.com/Abhilasha-Mangal/AutoHRise.git
cd AutoHRise

Installation

Before diving into the app, ensure that your environment is set up with all the necessary dependencies:

pip install -r requirements.txt

Enviourment Setup

To setup the eniourment please follow the below steps:

1. Use the sample file 'app/.env.example' to set up your Watsonx.ai and WatsonX Discovery credentials.
2. Enter all required credentials in the file.
3. Rename the file to app/.env after adding the credentials.

AutoHRise App

For a visual analysis of app, launch the streamlit dashboard:

 cd app
 streamlit run autohrise_app.py --server.port 8502 --server.fileWatcherType none

Future Scope

1. Connect the Flask API with WatsonX Assistant or WatsonX Orchestrate for a chat interface.
2. Real time resume screening from public directory.
3. Automatic candidate evalaution

Team

• Abhilasha Mangal
• Jatindra Suthar
• Sagar Padwal
• Bhargav Mankad