top of page

How Generative AI is transforming traditional analysis tools

Generative AI Bard Chat GPT CoPilot
Photo by Mojahid Mottakin on Unsplash

Everyone is talking about Chat GPT now!

It is through Chat GPT that even non-technical folks are accessing AI at a never-before scale. And that brings us to Generative AI. Generative Artificial Intelligence is a hypernym for different models capable of generating text, images, or other media in response to prompts and Chat GPT is just one of the many Generative AI applications. How exactly is Generative AI interacting with the data science and software development environment right now? Let us get into some specific areas to understand more about the topic and learn how we can make use of this technology in day-to-day work now and in the near future.

Before that, let me share a brief introduction to Generative Artificial intelligence technology. What exactly it is and are they and how do they work? Generative AI models are developed by training them with training data using neural network machine learning techniques. During the training process, a model learns the classifications, characteristics, and patterns of a large number of training inputs based on labels provided by researchers. Once trained, Generative AI can then generate new output against prompts or inputs. Different kinds of Generative AI algorithms produce different outputs like Chat GPT, Bing Chat and Bard (Google’s chatbot) produce text and Github Copilot generates codes. Imagen, DALL-E and Adobe Firefly are some of the text-to-image generators. Alphafold is an AI program trained on a sequence of amino acids which predicts protein structures as output.

The scope is so vast that in this blog I will just restrict the topic to Generative AI in some of the most commonly used analysis tools and software development platforms.

Excel and other Microsoft Office applications.

Microsoft 365 Copilot developed by Microsoft is an artificial intelligence assistant feature based on GPT-4 large language models (LLMs) for Microsoft 365 applications and services. It was announced by Microsoft in March 2023. The API used by Copilot is Microsoft Graph, which assesses the context and Microsoft 365 user data before adjusting and forwarding the user prompt to the LLM. Once the LLM responds, Microsoft Graph undertakes further context-specific processing before transmitting it to Microsoft 365 apps to produce the final content. We will keep coming back to Copilot at various points in this blog.

When working on Excel spreadsheets, Copilot is equipped to help users in various ways such as formatting data, generating PivotTables, creating graphs, identifying trends, and summarising information. Copilot can guide users by providing Excel commands and suggesting formulas to investigate user questions. In Word and PowerPoint, Copilot can produce first-draft Word documents or PowerPoint slides which the user can later tailor as required.

Another way of using Generative Ai is integration with Chat GPT. Although the ChatGPT Plugin for MS Excel is free to use on any version of MS Excel, it's important to note that the OpenAI API come with charges. The exact amount you'll be charged depends on your usage of the API. It is crucial to keep your API key confidential to prevent others from using it and incurring charges on your behalf.

Microsoft released Excel Labs in 2023, a project from Microsoft Garage. It is an add-in that allows users to test experimental ideas and provide feedback to help evolve Excel. The Garage is where Microsoft tests experimental projects to receive early feedback from customers and determine product market fit. With Excel Labs, users can expect to find early-stage ideas that Microsoft is exploring and wants to evaluate with customers. While some ideas may not make it to the Excel product, Microsoft believes that feedback from users is crucial to creating exciting features that transform what's possible in Excel.

The first two experimental ideas released in Excel Labs are - advanced formula environment which is an advanced editor for complex formulas. Another one is a custom function called LABS.GENERATIVEAI.

Tableau GPT and Tableau Pulse

It was recently announced at Tableau Conference 2023 that Tableau is working on a new tool called Tableau GPT and an allied tool called Pulse. Details of these services would be available soon.

Tableau GPT will make analysis and visualisation more accessible and easier for everyone. Collaborating with OpenAI, Salesforce has developed Einstien GPT, which is the foundation for Tableau GPT. By integrating generative AI into the user experience of Tableau, this assistant aims to help users work more effectively, learn more quickly, and communicate more efficiently. Simply put, Tableau GPT is an assistant that will take natural language queries as input and will return outputs based on the particular requirement, eg., if the user wants to create a calculated field, it can generate calculations that the user can copy and paste into the calculation window.

Tableau pulse can be a user’s personalized landing page with a newsfeed and users can personalize what they want to see as major highlights on this landing page – they can build metrics, KPIs, specific visualizations, time periods etc.

Power BI

We already talked about CoPilot. Power BI added Copilot mode in May 2023. Thanks to the latest AI assistant, users can obtain data, edit DAX calculations, and generate reports or data visualisations using voice commands. Copilot is also capable of answering data-related questions in a conversational manner and creating narrative summaries for reporting purposes.

Of course there’s Chat GPT integration option as well. With ChatGPT's assistance, users can easily create intricate calculations and queries for their Power BI models, identify and resolve errors, and improve report generation. Moreover, ChatGPT can provide extra support to users working with Power Query M functions. It can help troubleshoot any issues with these functions, generate data prep steps, and even write Power Query code using natural language instructions.

Azure Cognitive Services offers highly advanced AI algorithms for Power BI. These AI models are pre-trained and customisable, and they come packaged as APIs. With container deployment, Cognitive Services can provide cutting-edge analytical capabilities to a wide range of products on both cloud and edge platforms.

Within Power BI, it is possible to enhance pre-existing dataflows through the utilisation of Cognitive Services models that are readily available through a user-friendly graphical interface.


The potentials for Generative AI use cases in traditional analysis tools are tremendous. As we see the Copilots gearing up to improve productivity at the workplace, Generative AI can also be applied for analysis of business-specific use cases beyond just transforming analysis tools and software development. For example, Gen AI can be used for marketing to create content for email marketing campaigns including text and images. It can also act as a Concierge to content writers and aid with research and generating initial drafts. While we enter into a new era of the AI world, the need for a responsible approach to AI is also paramount. Protecting privacy, avoiding bias, and avoiding mistakes are crucial as we take steps to bring AI into our midst as digital ally.


A New Day for Data—Product Innovations from Tableau Conference 2023. (n.d.). Tableau.

Tableau Tim. (2023, May 11). Tableau Conference 2023 Keynote Roundup | Tableau Conference 2023 [Video]. YouTube.

How Tableau GPT and Tableau Pulse are reimagining the data experience. (n.d.). Tableau.

Spataro, J. (2023, May 16). Introducing Microsoft 365 Copilot – your copilot for work - The Official Microsoft Blog. The Official Microsoft Blog.

60 views0 comments

Recent Posts

See All


Rated 0 out of 5 stars.
No ratings yet

Add a rating
bottom of page