Photo by the editor
ChatGPT has become an OpenAI product that is changing the way the world works. Many readers are already using them or at least testing them. Even though it helps us, I don’t think we can go back to the venerable way of working.
One of the innovations provided by OpenAI is the GPT store, where users can develop their own GPT models and share them publicly. There are over 3 million ChatGPT custom GPTs open from the creators. Indeed, some of them may be useful to improve the activities of data scientists.
This article discusses 7 GPTs from the GPT Store that can improve your Data Science workflow. What are these GPTs? Let’s get on with it.
As a side note, I would exploit Telecom Churn dataset from Kaggle as a sample dataset to be used by GPT.
Let’s start with what the ChatGPT teams have created for us, i.e Data analyst. This is a custom GPT, specially trained to analyze our data and visualize it as needed. By dropping a file, such as a CSV file, and prompting us for what we need, Data Analyst GPT can perform the task automatically.
For example, I ask a Data Analyst to develop a churn correlation analysis based on a dataset I give him.
Data analyst performs correlation analysis (author’s photo)
You can request further analysis from a GPT data analyst. You can also exploit GPT to make all the code available for self-execution if necessary.
The next GPT we will discuss is GPT machine learning. This custom GPT is designed to be an assistant for all your machine learning and data analytics activities. Relevance includes discussing, learning, and developing appropriate algorithms for our data projects.
As an example, I ask Machine Learning GPT to develop a model based on our sample dataset to predict churn. Here’s the result.
Machine learning conducts model experiments (author’s image)
GPT can provide an excellent comparison of used models. If we continue, we can ask the model to iterate with more models, perform hyperparameter tuning, and ask GPT to provide reasons for each action.
As in the previous entry, GPT Machine Learning Engineer provides users with an assistant in developing a machine learning model. You can place your dataset and ask GPT to provide the necessary steps and full code.
What sets a machine learning engineer apart is that his or her GPT determines the design of models to automate convoluted tasks, especially when deploying models. GPT is a good place to discuss model structure and how to deploy the model to production.
Speaking of structuring models, GPT is also suitable for helping us structure our code for machine learning modeling. One of the best coding assistants I have found is AutoExpert. This is GPT that is designed to lend a hand you as a steadfast developer assistant.
GPT was developed with additional code generation capability, online access to the latest APIs, and custom commands to save session state that can be used in a later session if needed.
Using this GPT will lend a hand you generate convoluted code for any purposes needed in your data science business. It also provides a code structure and script to lend a hand you execute them better.
Let’s move on from the technical coding part and move on to the theoretical part. As we know, data science work is about continuous learning, especially when dealing with novel exploit cases. With the growing amount of research in the field of data science, it is sometimes challenging to find the perfect research that could lend a hand our exploit cases. This is where ScholarGPT comes in.
This GPT will lend a hand you find the latest research article related to our exploit cases. With a basic prompt it would give us a selection of the latest articles related to the problem we want to solve.
For example, the text below is the result of a study by ScholarGPT, to which I submitted our dataset and asked them to provide me with a research article on churn prediction.
Title: “Transparency in Decision Making: The Role of Explainable Artificial Intelligence (XAI) in Customer Churn Analysis”
- Authors: C ÖZKURT
- Year: 2024
- Summary: This study focuses on predicting customer churn and explaining the causes of customer churn using machine learning, in particular on the detailed analysis of customer churn in the telecommunications sector.
- To combine: Read the newspaper?source?.
ScholarGPT provides many more research articles to choose from, so you can choose which one applies to your exploit cases.
The next GPT we will discuss is Whimsical diagram. Many data science activities are not always about model research and development. Many times we need to visualize our workflow and explain what our work will look like. This is where Fancy GPT Diagrams will lend a hand you.
This GPT aims to explain and visualize concepts using flowcharts, mind maps and sequence diagrams. Providing prompts and a data source can lend a hand us provide visualizations that make our work easier.
For example, I asked the model to provide me with a suggestion diagram from the Churn dataset that suggested a feature-based churn visualization. Below is the image result.
Opting Out Due to Features (Image Generated by Whimsical Diagram GPT)
You can further discuss with GPT to find the perfect workflow diagram for your data science work.
The last one is Canva GPT, which can lend a hand us report our results. As we know, Canva is a service platform that helps you design everything from logos to profile photos, banners, and presentations. With Canva, GPTs can lend a hand us get the best design for our analysis.
Data science is about communicating our results to others, so it is significant to have reliable results that are presented in a way that the audience understands. With Canva GPT, we can ask for suggestions for a suitable design. For example, I asked a model to provide me with a design that would be perfect for showing off churn statistics.
Choosing an churn statistics design (Canva GPT)
GPT would give us design options and we could choose which ones we prefer or give additional hints to get different designs.
This article discusses seven custom GPTs available in the GPT Store that could improve our data analysis workflow; They are:
- Data analyst at ChatGPT
- Machine Learning by Maryam Eskandari
- Machine Learning Engineer at Hustle Playground
- AutoExpert (programmer) Author: llmimagineers.com
- ScholarGPT by awesomegpts.ai
- Fancy diagrams at whimsical.com
- Canva at canva.com
I hope it will lend a hand! Do you have any suggestions for GPT that should be on this list? Please include them in the comments as well.
Cornelius Yudha Vijaya is an assistant data analytics manager and data writer. Working full time at Allianz Indonesia, he loves sharing Python tips and data through social media and writing media. Cornellius writes on a variety of topics related to artificial intelligence and machine learning.