Worksheet software concept with illustrations in green and blue colors. Can be used for animations etc. [+] Banners and presentations.
Getty
As part of a series, I try to put theory into practice and help readers of this column apply AI in their daily work. In my first article, I explained how to get started with text-to-speech technology in a few easy steps. Today, I’ll talk about an increasingly popular use case (although not without its drawbacks): data analysis.
How does AI work in spreadsheets?
You’re probably used to having a data analyst on your team who is a spreadsheet whiz and the go-to person when you need some custom Excel work. I used to be like that too, using a combination of tools like Excel, MS Access (with SQL), SAS, and more to do large-scale data analysis. But now, thanks to AI, the process is much simpler. A talented analyst can have a much bigger impact because they no longer have to waste time on pointless tasks like data cleansing.
Here’s a rough outline of how AI works.
Data analysis: AI deciphers the structure of spreadsheets and identifies rows, columns, headers, and data types. Pattern recognition: Using machine learning, AI recognizes patterns and correlations in the data, including those that may not be immediately apparent to a human analyst. Data cleaning: Automatically detects and corrects errors and inconsistencies, such as missing values and outliers. Insight generation: By applying statistical models, AI derives insights, predicts trends, and makes predictions based on the data.
This means that a good analyst who knows how to ask the right questions will generate more insights than an analyst without AI.
How can I try this?
Create a Chat GPT Plus account. There are other options available here, but most of you will be familiar with Chat GPT, so stick with the options you are familiar with.[Explore GPTs]Find the tab. Once you get there, select Data Analyst GPT (image) and a chat interface will open. Select the dataset you want to use and download it as a CSV (or any other format). If you’re looking for example data to use, head over to Kaggle – a great repository with lots of interesting stuff. Upload your CSV to the chat window and tell Chat GPT to work with your spreadsheet. I often use a prompt like “Read this CSV I’m about to upload, remove any errors, and prepare for a series of questions I’ll ask you for further analysis.” Ask questions. This is key. Ask good questions. Examples include “Uncover 5 unobvious insights from this data” or “Help me find long-term trends that I’ve missed for some reason.”
In the example below, I downloaded a dataset about New York real estate from Kaggle. I then uploaded it to Chat GPT and asked it to reveal some insights. Here are the results I got:
Example Chat GPT output from a data analysis job.
Sunil Rajaraman
Of course, the correlation analysis is not as deep, but it was much more insightful and very quickly, without any additional prompting.One obvious drawback I noticed is that as the dataset gets larger, both the speed and accuracy of the analysis can become problematic.
How should data analytics be used today?
Today, there are many practical use cases for data analytics that you should consider using, including:
Financial forecasting – Ingest existing financial information to see if trends like seasonality can be more easily identified with AI. Improve HR practices – Reduce employee numbers, turnover, retention, and improve recruiting processes by understanding what’s wrong. Competitive analysis – Evaluate competitors’ SEO strategies, etc., and keep track of your competition.
The options are endless, but you’ll be missing out if you don’t try out some of these tools sooner rather than later – they’re only going to get better by leaps and bounds.