How To Use Artificial Intelligence Today: Advanced Data Analysis

As part of an ongoing series, I am trying to take theory into action and help anyone reading this column apply AI to their day-to-day work lives. In my first piece, I covered how you can get started with text-to-speech technology in a few simple steps. Today, I’m going to cover data analysis, which is going to be an increasingly popular use case (which is not without its flaws).

How does AI work with spreadsheets?

You are probably accustomed to having a data analyst on your team who is an absolute spreadsheet whiz, and is your go to person when you need some custom Excel work done. I was once that person – I used a combination of Excel, MS Access (with SQL) and tools like SAS to get large data analysis done. But now, with AI, the process is much simpler – your best analyst can have a significant amount more leverage because they won’t need to waste their time on useless tasks like data cleansing.

Here is a rough idea of how the AI works:

1. Data Parsing: AI deciphers the structure of your spreadsheet, identifying rows, columns, headers, and data types.
2. Pattern Recognition: Using machine learning, the AI recognizes patterns and correlations within the data, even those not immediately apparent to human analysts.
3. Data Cleaning: It automatically detects and corrects errors or inconsistencies, like missing values or outliers.
4. Insight Generation: By applying statistical models, AI derives insights, forecasts trends, and makes predictions based on the data.

This is all to say, a great analyst who knows how to ask the right questions will generate more insight than an analyst without AI.

How can I try this out?

1. Create a Chat GPT Plus Account. There are other options here available to you, but since most of you are probably familiar with Chat GPT – go with what is familiar.
2. Search the ‘Explore GPTs’ tab on the left hand pane. Once you are there, select the data analyst GPT – pictured – and you’ll open a chat interface with it.
3. Select which dataset you want to work with, download as a CSV (or really any other format). If you’re looking for sample data to work with – go to Kaggle for a great repository of lots of interesting things.
4. Upload the CSV to the chat window, and prompt Chat GPT to interact with the spreadsheet. I often use the prompt – ‘Please read the CSV I’m about to upload, clean it of any errors, and please prepare yourself for a series of questions that I’ll ask for deep analysis.’
5. Ask questions. This is the key – ask good questions. Examples might include ‘Uncover 5 non-obvious insights from this data’. ‘Help me spot trends that I’d somehow miss that are long term’.

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 uncover some insights – here is what it came up with:

Of course correlation analyses aren’t very deep, but without much additional prompting it got a lot more insightful very, very quickly. One obvious shortcoming that I’ve noticed, as a disclosure, is that as datasets get larger it sometimes can struggle with both speed and accuracy of analysis.

How should you be using Data Analysis today?

There are a number of practical use cases for data analysis today that you should consider using – examples include:

1. Financial forecasting – ingest your existing financials to see if there are trends like seasonality you can more easily identify with AI.
2. Better HR practices – reduce employee, churn, retention and make your recruiting processes better by figuring out what’s going wrong.
3. Competitive analysis – track competition by assessing your rivals’ SEO strategy, etc.

The options are kind of limitless, but it would be a huge miss if you did not start playing with some of these tools earlier. They are only going to get exponentially better from here.