Project Documentation of VOC Dashboard

1. Objective To create a Streamlit application that fetches sentiment and topic data from an MSSQL database, generates insights using OpenAI's GPT-4 model, and displays interactive visualizations and PowerBI reports.

2. Workflow a) Setup Streamlit Environment: Configure the Streamlit application with a wide layout and add a company logo.

   b) Fetch Data from MSSQL: Connect to the MSSQL database, execute SQL queries to retrieve data from the Sentiment_Data and Topic_data tables, and store the data in Pandas DataFrames.

   c) Generate AI Insights: Use the OpenAI API to analyze the fetched data and generate detailed insights.

   d) Display PowerBI Reports: Embed a PowerBI report in the Streamlit application.

   e) Dynamic Data Interaction: Allow users to interact with the data and generate dynamic insights and visualizations based on user interactions.

3. Requirements

Software and Libraries:

Streamlit
Pandas
Plotly
SQLAlchemy
OpenAI API
Streamlit Components
PowerBI

Database: MSSQL Server with tables Sentiment_Data and Topic_data

APIs and Keys: OpenAI API Key

Configuration: Database connection parameters (server, database, username, password, driver)

4. Solution: 1_📈_Report and Insights.py

Project Documentation of VOC Bot

1. Objective To develop a Streamlit application named VOC Bot that allows users to interact with sentiment and topic data from an MSSQL database, generating insights and summaries using OpenAI's GPT-4 model, and presenting results through interactive chat and visualizations.

2. Workflow a) Setup Streamlit Environment: Configure the Streamlit application layout, add a custom title, and display the company logo.

   b)Database Connection: Establish a connection to the MSSQL database and reflect the schema using SQLAlchemy.

   c) Generate AI Insights: Formulate a prompt for GPT-4, generate SQL queries, and retrieve summarized insights based on user questions.

   d) User Interaction and Chat: Implement a chat interface that takes user inputs, processes them to generate responses, and displays both user and assistant messages.

   e) Display Results and Error Handling: Generate dynamic responses and handle errors gracefully.

3. Requirements Software and Libraries:

   Streamlit Pandas SQLAlchemy OpenAI LlamaIndex Streamlit Components PyODBC

   Database: MSSQL Server with tables Sentiment_Data and Topic_data

   APIs and Keys: OpenAI API Key

   Configuration: Database connection parameters (server, database, username, password, driver)

4. Solution:

   2_🤖_VOC bot.py