TalentScout Chatbot

TalentScout Chatbot is a recruitment assistant built using Streamlit and LangChain. It guides candidates through a recruitment screening process, collecting their information, conducting a technical interview, and exporting the data for further evaluation. The chatbot also integrates with MongoDB to store interview data securely.

---

Live Demo

You can access the live version of the TalentScout Chatbot here:

👉 TalentScout Chatbot Live

---

Project Overview

TalentScout Chatbot is designed to streamline the recruitment process by automating candidate screening. It collects candidate information, conducts technical interviews, tracks progress, and stores the data in a MongoDB database for further evaluation. The chatbot uses LangChain for prompt generation and Streamlit for the user interface.

---

Features

• Candidate Information Collection: Collects personal details like name, email, phone, experience, and tech stack.
• Technical Interview: Asks technical questions based on the candidate's tech stack.
• Progress Tracking: Displays interview progress in real-time.
• Data Export: Allows exporting interview data in JSON format.
• MongoDB Integration: Automatically saves interview data to a MongoDB database.
• Privacy Notice: Displays a data handling policy to ensure transparency.

---

Installation Instructions

1. Clone the Repository:

   git clone https://github.com/your-repo/talentscout-chatbot.git
   cd talentscout-chatbot

2. Install Dependencies: Ensure you have Python installed. Install the required packages:

   pip install -r requirements.txt

3. Set Environment Variables: Create a .env file in the root directory and add your Groq API key and MongoDB connection string:

   GROQ_API_KEY=your_groq_api_key
   MONGO_URI=mongodb+srv://<username>:<password>@cluster0.mongodb.net/<database>?retryWrites=true&w=majority
   

4. Run the App: Start the Streamlit app:

   streamlit run app.py

5. Access the App: Open the app in your browser at http://localhost:8501.

---

Usage Guide

1. Start the Application:

   • Launch the app locally or access the live demo link.
   • Follow the chatbot's instructions to provide your information.

2. Complete the Interview:

   • Answer the technical questions based on your tech stack.
   • Track your progress using the progress bar.

3. Export Data:

   • After completing the interview, the data is automatically saved to MongoDB.
   • You can also download the data in JSON format.

---

Technical Details

Libraries Used

• Streamlit: For building the interactive web application.
• LangChain: For generating dynamic prompts and interacting with the LLM (ChatGroq).
• pymongo: For integrating with MongoDB.
• dotenv: For managing environment variables.

Model Details

• LLM: ChatGroq (model: llama-3.3-70b-versatile / meta-llama/llama-prompt-guard-2-86m) is used for prompt generation and information extraction.

Architectural Decisions

• Modular Design: The code is divided into utility files for better maintainability.
• Session State: Used to manage the state of the application, including candidate data and interview progress.
• MongoDB Integration: Ensures secure and scalable storage of interview data.

---

Prompt Design

Information Gathering

Prompts are crafted to collect specific candidate details, such as name, email, phone, experience, and tech stack. For example:

• "What's your full name?"
• "What position are you applying for?"

Technical Question Generation

Prompts dynamically generate technical questions based on the candidate's tech stack. For example:

• "Generate a technical question for a candidate proficient in Python and Django."

Relevance Check

Prompts ensure user messages are relevant to the recruitment process. Irrelevant messages are politely redirected.

---

Challenges & Solutions

Challenge 1: Handling Irrelevant User Input

• Problem: Users might provide irrelevant input during the interview.
• Solution: Implemented a relevance check using LangChain prompts to filter irrelevant messages and redirect users.

Challenge 2: MongoDB Integration

• Problem: MongoDB's ObjectId is not JSON serializable.
• Solution: Converted ObjectId to a string before exporting data.

Challenge 3: Dynamic Prompt Generation

• Problem: Generating diverse and meaningful technical questions.
• Solution: Used LangChain's LLM capabilities to craft prompts based on the candidate's tech stack.

---

Directory Structure

talentscout-chatbot/
│
├── app.py                     # Main entry point for the Streamlit app
├── requirements.txt           # Python dependencies
├── README.md                  # Project documentation
├── utils/                     # Utility modules for modularity
│   ├── __init__.py            # Makes the folder a Python package
│   ├── session_state.py       # Functions for initializing and managing session state
│   ├── prompts.py             # Functions for generating prompts and messages
│   ├── validation.py          # Functions for validating user inputs
│   ├── llm.py                 # Functions for interacting with the LLM (ChatGroq)
│   ├── progress.py            # Functions for calculating and rendering progress
│   ├── export.py              # Functions for exporting candidate data
│   └── sidebar.py             # Functions for rendering the sidebar

---

Future Improvements

• Add support for multiple languages.
• Enhance the technical question generation with more diverse question types.
• Integrate authentication for HR personnel to access candidate data.
• Implement persistent session storage for long-running interviews.

---

Contributors

• Arvinder Singh Dhoul: Developer and Maintainer