How to Automate Customer Sentiment Analysis Reports: A Guide for Developers

How to Automate Customer Sentiment Analysis Reports: A Guide for Developers

If you’re a developer, especially for an automated platform company, this guide is for you. It walks you through a script we created to generate customer sentiment analysis reports automatically. Creating these reports manually takes a lot of time and energy — something few developers have. However, you can automate the report generation process with AI and some scripting. You can find links to download the script file and related materials at the bottom of this guide.

What you’ll need to run the script

  • Customer Reviews Data — You should obtain customer reviews data from a reliable source. For this guide, we’re getting the data from the eCommerce Reviews API. It provides product information and customer reviews from 900+ eCommerce and marketplace sources.
    • You need an API key to use the eCommerce Reviews API, and you can get one by contacting This guide includes a free NDJSON file with sample reviews data if you would like to experiment with the script without using the API.
  • OpenAI API — You’ll use OpenAI’s API to leverage the GPT-4 and DALL·E models. GPT-4 analyzes and summarizes the text from customer reviews, while DALL·E generates a main image for the report.
  • Python — We’re using Python to automate the report creation process. You’ll need to ensure you can run Python code on your machine.  
How and ChatGPT automate your report

Automating customer sentiment analysis reports: script breakdown

The script fetches product reviews from an external NDJSON file generated by the eCommerce Reviews API. Next, the script calls the OpenAI API, using its text and image models to analyze the reviews for positive and negative sentiments. It then compiles these findings into a structured report, outputting them into a Word document. The document contains the completed customer sentiment analysis report. 

Here is the detailed breakdown of the script:

Import files, packages, and modules

First, the script imports the NDJSON file, Python packages and modules, the OpenAI Python API library, and other necessary files.

Set global variable and access API key

Next, the script accesses the Open AI API key through the development environment. It also includes a global variable where you can set the number of reviews included in the report.

Orchestrate entire process (main)

Towards the end of the script, you’ll see the “main” function. It orchestrates the entire process — from reading the reviews and generating a report main image to generating the report text and creating the final Word document.

Define functions

Now we define the different functions of our script:

Read NDJSON file (read_ndjson_file)

Reads a NDJSON file and returns its content.

Send prompt (call_gpt_completion)

Sends a prompt to the GPT-4 model and receives a response.

Extract points (extract_points)

Extracts key points from reviews based on sentiment (positive/negative).

Generate title (generate_title)

Creates a title for the sentiment analysis report.

Generate introduction (generate_intro)

Generates an introductory paragraph for the report.

Generate negative report (create_negative_report)

Creates a report based on negative feedback.

Generate positive report (create_positive_report)

Creates a report based on positive feedback.

HTML to formatted text (html_to_word)

Converts HTML content to formatted text in a Word document. It handles bold text and bullet lists.

Add hyperlink (add_hyperlink)

Inserts a hyperlink into a Word document paragraph.

Add title placeholder (insert_titles_in_text)

Adds a placeholder for inserting the titles.

Generate image (generate_article_image)

Generates a main image for the report using OpenAI’s DALL-E model with a specific prompt.

Download image (add_image_from_base64)

Downloads an image from a URL and adds it to a Word document.

Create Word doc (create_word_doc)

Assembles the various components into a formatted Word document.

AI + Python = a powerful automation tool

This script demonstrates an advanced use case of integrating AI-powered content generation with document automation in Python. It’s a comprehensive example of how combining various Python libraries with AI models can produce a powerful automation tool.

Download the example code and files:

  • The full Python script

To run the script:

  • Ensure that Python and the required Python libraries are installed on your machine.=
  • Set your OpenAI API key in your development environment.
  • Place the NDJSON file with the product and review information in the specified directory.
  • Run the script.

Ready to automate customer sentiment analysis for your organization? Talk to one of our experts today.

Spread the News

Not subscribed to our Dark Web Pulse updates?

By submitting you agree to's Privacy Policy and further marketing communications.

Feed Your Machines the Data They Need

Feed Your Machines the Data They Need

Subscribe to our newsletter for more news and updates!

Ready to Explore Web Data at Scale?

Speak with a data expert to learn more about’s solutions
Create your API account and get instant access to millions of web sources