Generative AI for Data Analytics: All Your Questions Answered

Generative AI for Data Analytics

Phew! What a time to be alive for generative AI. 😮‍💨 Investors have been betting big on GenAI-focused start-ups, and millions of users are asking ChatGPT their probing questions every day. 

For the data and analytics world, it’s no secret that GenAI has the potential to massively transform the field of analytics by introducing new capabilities, enhancing existing processes, and enabling new applications. Basically, generative AI is empowering more people at every level of the organization to make more data-driven decisions—i.e., self-service analytics.

But let’s peel back the layers a bit and take things piece by piece. After all, it’s only been a little over a year since GenAI first emerged into the world as ChatGPT, so naturally, many questions remain about introducing generative models into analysis workflows.

Here are some of the most commonly asked questions we’ve seen surrounding generative AI for data analytics. We even asked ChatGPT for some advice on what’s important. (It’d be wrong not to. 💁‍♀️)

Starting with the basics, what is generative AI?

A type of artificial intelligence, generative AI creates text, images, or other media by using a generative model, which refers to a large language model (LLM): e.g., Google’s PaLM, Meta’s LLaMa, and the most commonly known LLM, OpenAI’s GPT. LLMs—which are trained to learn patterns from a massive amount of data—analyze a text prompt to make sense of what you’re asking and then put their training to work to generate a contextually relevant response.

Why are businesses using generative AI for analytics?

GenAI brings so much to the table for the development of data analytics. We’ll get into more specific use cases, but here are three main benefits at a high level:

Democratizing data access: Generative AI models with natural language processing capabilities like GPT are transforming data into valuable insights for users across the board. Non-technical stakeholders are able to easily leverage data in meaningful ways by simply opening their analytics tool, asking a question, and receiving an answer that explains the key insight.

Saving businesses significant time and resources: When organizations can expand self-service data exploration to more users, they can cut down on time-consuming analysis work, reduce what’s often an overwhelming BI backlog, and free up time for IT and data engineering team members. For data analysts, GenAI-based narratives can help simplify the data analysis process and, in turn, enable them to spend more time asking new and more meaningful questions of their data. In addition, GenAI’s natural language-to-code capabilities provide another way to free up resources.

Driving better business outcomes: Overall, generative AI has the ability to significantly transform data and analytics for organizations by helping them extract valuable insights from their data; personalizing customer experiences by better predicting customer behavior; and enabling easier and quicker data-driven decision-making to optimize business operations and, ultimately, drive revenue.

For more, check out 5 Ways ChatGPT and LLMs Could Influence BI/Analytics.


What are some common generative AI analytics use cases?

Across industries, here’s how generative AI is more specifically being used to enhance data analytics:

Analytics chatbots: Many analytics platforms are integrating GPT so users can conversationally ask questions of their data and get an answer in natural language (or a recommended visualization, chart, or graph). Some GPT chatbots can also produce narrative summaries, build reports, or, as previously mentioned, translate natural language to code.

Analytics workflow augmentation: GenAI can generate synthetic data that helps train machine learning models and complements existing datasets, providing benefits across the organization: 

  • For the business user: greater context and narrative summaries
  • For the analyst: metadata enrichment
  • For the data scientist: automated code generation and validation

Data prep and cleansing: When integrated with an analytics platform, GenAI can augment data preparation through automated data classification by identifying relevant features from raw data. It can also automatically generate metadata, adding, for instance, synonyms and descriptions of columns, to improve the data’s accessibility and usability. Generative AI can also produce a formula just from a user’s description with natural language, eliminating the need for the user to memorize a formula from a spreadsheet.

Predictive analytics: Generative AI for predictive analytics can provide valuable insights for planning and decision-making. Generative models, especially ones that can handle sequential data, can be applied to time series prediction and forecasting. In addition, generative AI can enhance the training of a predictive model, enabling more accurate and robust predictions for organizations across industries: e.g., life science and pharmaceuticals, CPG, healthcare, finance, marketing—you name it.

Just how big is generative AI in analytics really getting?

It’s estimated that ChatGPT now has more than 100 million weekly active users, and OpenAI is reportedly shelling out $700K/day to operate it. (Of course, you can’t actually ask ChatGPT for this information—it doesn’t have real-time data, and it can only provide information based on its “last knowledge update.” But we will get more into common GenAI analytics cautions shortly. ⚠️)

So, based on the massive popularity of GPT, it makes sense that many BI and analytics vendors are focusing on conversational analytics within their platforms, introducing GPT-based chatbots as interfaces for search and narrative summaries.  

Overall, generative AI is helping extend the power of analytics across an organization. The application of LLM technology on a platform built for multi-persona analytics in particular (data scientists, analysts, and business users) enables more opportunities for collaboration among an organization.

To learn what’s really moving the analytics needle versus what’s just hyped, check out our in-depth roundup of GenAI for BI and analytics vendors


What are some common concerns about generative AI in data analytics?

Incorrect or out-of-date information: The reality is that LLMs can hallucinate—i.e., generate information that’s either incorrect or nonsensical—leading to inaccuracies or misinterpretations in analysis. Businesses should also take into account the fact that an LLM-based chatbot might not have the right context for its narrative analysis if it wasn’t trained on data from a certain time period.

Difficult integration: Although generative AI data analytics tools are becoming increasingly more accessible, it’s not a simple plug-and-play solution.

An organization’s data and domain experts must still exercise caution, collaboration, and patience when adopting the technology, which requires setting up, mapping, and fine-tuning to specific needs (e.g., use cases, data size, individual expertise).

Other practical implications: While it’s clear that the combination of generative AI and analytics offers transformative potential, it’ll probably take some time before businesses see the technology incorporated into their analytics workflows on a daily basis on an enterprise level, rather than in piecemeal fashion.

Organizations still need to keep certain things in mind:

  • Proper data security: GenAI-integrated analytics platforms must have robust data encryption, access controls, and compliance certifications to protect sensitive data.
  • Long-term costs: With so many companies incorporating GPT into their analytics platform, the costs associated with using generative AI tools could add up. Thus, it’s important for businesses to do their due diligence when investing in the right technology.
  • The need for human intervention: Don’t expect GenAI to completely replace the work of AI analytics experts. (It’s not a magic bullet!) While generative AI is opening groundbreaking new doors for self-service data exploration and analytics workflow enhancement, the data experts still have their vital place in ensuring data quality, defining metrics, identifying analysis pitfalls, interpreting results, and making well-informed strategic decisions.

Read more about tackling problems with the current approach to generative AI for analytics: Rethinking GenAI and Business Intelligence.


Learn more about GenAI’s data and analytics transformation in 2024

2023 set the stage for a transformative 2024 for the world of GenAI and analytics. Businesses are poised to integrate the technology with their data (if they haven’t done so already), revolutionize their analytics workflows, and boost employee productivity across the organization.

Watch our recent webinar where we answered this question: How can this groundbreaking technology truly be harnessed to unlock unprecedented levels of data consumption and productivity within an organization?

genai data analytics



Read Similar Posts

  • 7 Takeaways from the Gartner Data & Analytics Summit 2024
    Business Intelligence

    7 Takeaways from the Gartner Data & Analytics Summit 2024

    The 2024 Gartner Data & Analytics Summit was a jam-packed three days of sessions and networking opportunities for data & analytics leaders. Here's what Tellius took away from the event.

  • 5 Common Pitfalls to Avoid When Launching a Self-Service Analytics Program
    Business Intelligence

    5 Common Pitfalls to Avoid When Launching a Self-Service Analytics Program

    Here are some common pitfalls we've seen for organizations launching a self-serve analytics vision—and how to avoid them to maximize your odds of success.

  • GenAI for BI & Analytics Roundup
    Business Intelligence

    GenAI for BI & Analytics Roundup

    Here's a roundup of GenAI offerings we’ve seen announced or launched in the business intelligence and broader analytics market.