The Real-Time AI-Powered Sales Intelligence Tool is designed to revolutionize live sales calls by providing actionable insights and suggestions to sales teams in real-time. This cutting-edge tool leverages advanced AI models for sentiment and intent analysis, integrates with CRM data and Google Sheets, and provides optimized negotiation strategies to enhance customer engagement and drive sales outcomes.
Real-time Speech Recognition for continuous audio recording from the buyer.
Automated Speech-to-Text Transcription to convert recorded audio into text.
Sentiment and Intent Analysis powered by state-of-the-art NLP models.
Recommending Laptops based on the given input.
Intelligent Negotiation Terms Generation based on buyer sentiment and intent.
Seamless Google Sheets Integration to record and manage buyer interactions.
Fully functional Workflow Integration to streamline processes end-to-end.
The project workflow includes 8 major steps:
Tools Used: PyAudio, SpeechRecognition libraries.
Functionality: Continuously records audio input from the buyer during live sales calls.
API: Google Speech-to-Text API.
Implementation:Enabled Google Drive and Google Sheets APIs.
Downloaded credentials.json file using a service account on Google Console.
Output: Converts audio into text for further processing.
Models Used:Sentiment Analysis: cardiffnlp/twitter-roberta-base-sentiment from Hugging Face.
Intent Analysis: facebook/bart-large-mnli from Hugging Face.
Objective: Understand the buyer's mood and intent to provide actionable insights.
Models used: Command-xlarge-nightly model from Cohere LLM
Approach: Suggest laptops based on the input given by the buyer and extract matched laptops names in the input to the product name in the dataset, recommends to the buyer.
Models used: Command-xlarge-nightly model from Cohere LLM
Approach: Extracts keywords based on sentiment and intent analysis.
Generates basic negotiation terms tailored to the buyer's context.
Aprroach: Summarizes the whole conversation and finalises the deal status based on the sentiment.
Model used: llama 3.3 70b versatile model from GROQ LLM.
Purpose: Records all buyer interactions and contextual data for tracking and analysis.
Implementation:Used the spreadsheet ID of a shared Google Sheet linked with the service account in the credentials file.
Objective: Seamlessly integrates all steps into a unified, functional workflow for real-time operation.
We welcome contributions to enhance this project. Please follow the guidelines below:
Check existing issues before creating a new one.
Provide a clear and concise description of the problem. Include steps to reproduce the issue, if applicable.
Fork the Repository: Create your own fork of the repository.
Clone the Repository: Clone the forked repository to your local machine.
git clone https://github.com/Joshithach18/InfosysSpringboard5.0-Group1.git
Create a Branch: Create a new branch for your changes.
git checkout -b feature/your-feature-name
Commit Changes: Commit your changes with a descriptive message.
git commit -m "Add description of your changes"
Push Changes: Push your changes to your forked repository.
git push origin feature/your-feature-name
Create a Pull Request: Submit a pull request to the main repository.
Follow PEP 8 for Python code.
Include comments and documentation for clarity.
Write unit tests where applicable.
Enhance existing features.
Optimize performance.
Integrate additional NLP or AI models for improved accuracy.
By contributing, you agree that your contributions will be licensed under the same license as the project.
Clone the repository:
git clone https://github.com/Joshithach18/InfosysSpringboard5.0-Group1.git
cd InfosysSpringboard5.0-Group1
pip install -r requirements.txt
Run the project:
Streamlit run frontend.py
This project uses:
Hugging Face Transformers
Cohere LLM
Groq LLM
Google Speech-to-Text API
Google Sheets API