Start using AI to analyze data. Here are the best tools to get started.
Friendly Reminder: AI is not going to replace you. Someone using AI might.
How do we prevent that? Become good at using AI tools.
How do we become good at AI tools? Same way we get good at anything.
Today, I’ll teach you what I learned analyzing 100,000+ data points with two popular data tools: ChatGPT & Julius AI.
Data Analysis with ChatGPT
You’re obviously probably familiar with ChatGPT. It’s the most famous AI tool. I used ChatGPT yesterday to help install a new security camera, create a cartoon (for this newsletter - stay tuned), and analyze data - all within 30 minutes
As of today, there’s two main ways to “do” data analytics on ChatGPT. One is to ask ChatGPT what to do and how to do it, and then execute outside the ChatGPT platform.
For example, “based on this data set, write an Excel formula that returns the average of column F”. ChatGPT would then give you a formula, you’d copy it, and paste it over to your local Excel file.
The other method is what used to be called the “Advanced Data Analysis tool” but now is just clumsily unnamed and magically triggered if you prompt ChatGPT the right way. Here, ChatGPT does the analysis for you. No need to copy and paste anything - the answers will be computed & shown to you inside the ChatGPT platform.
In this case, ChatGPT creates Python scripts that analyze your data, executes the scripts in a secure environment, and shows you the results.
Data Analysis with Julius AI
You may not have heard about Julius AI before - that’s ok. They’re new & upcoming. It’s not a generic AI tool - it’s specifically for data analytics. I don’t use Julius AI to create my half IRONMAN training plan, or to help me pick out my next camera lens (btw, new camera gets here in a week - I’m so excited to make more YouTube videos!). But I do use it to do data analysis.
Julius is a great data analysis AI tool. Simply upload your data set, ask questions in plain English, and get insights in seconds. It’s basically the equivalent of ChatGPT’s “Advanced Data Analysis tool” but faster and smarter. How?
Well, Julius is a platform built for only data analytics. ChatGPT is built for 1,000+ different use cases. It’s like the difference between a hatchet and a swiss army knife to cut a tree down. Could you do it with a swiss army knife? Probably - but it’ll take you much longer than just using a tool designed for the job.
Not only is Julius a chat-based tool, but it also has built in notebooks similar to Jupyter Notebooks. It has a click-and-drag graph maker & data table explorer. Next, when using AI, you need to be able to trust the answers. Julius does a great job explaining what it’s doing in each step in simple language. It also has better data visibility (it’s easier to see the created code & tables).
The compute ability on Julius is much better than ChatGPT. For starters, you can ramp up your RAM and supercharge your analysis speed. You can also install complex Python packages on Julius (something you can’t do on ChatGPT).
Recommended by LinkedIn
And most importantly, it’s just a little bit smarter as an AI analyst. It suggests things you might want to do next, it goes the extra mile in analysis, and has more customized settings & presets.
Simply, it’s an AI tool built for data analysis. You should check it out - you can get started for free here.
A Simple Challenge For You
It’s one thing to read about how to analyze data with AI and another thing to see it. I recently recorded a video that show cases the differences between ChatGPT and Julius AI. You can watch that here.
But what’s better than seeing? Experiencing. Here’s my 5-step challenge for you: (takes 30 minutes)
I promise, you’ll learn a lot!
The better you get at using AI to analyze data, the brighter your future as a data analyst
AI is a tool, not a threat.
Let’s learn it together!
p.s. I partnered with Julius AI to bring you this newsletter issue, but I genuinely enjoy using the tool and think you will too!
This image was created by ChatGPT and as you can see, AI is not perfect
Want to see how I analyze data with ChatGPT and Julius? Here's a full demo 👇
Did you tried Elkar ? https://elkar.co/ :)
Avery I followed your adviced and have jumped in to working with Julius and... Wow!! I (mostly) vibe-coded a fully refreshable dashboard on Energy Sources from government and other databases in 40 minutes...which is probably on the long side since I tend toward the cautious. One awesome use-case popped up right away when I prompted for data cleaning and after several minutes and multiple attempted fixes, Julius stated the dataset couldn't be used for analysis due to formatting issues (no time to look into it further at the moment) ; I had the freedom to use a different dataset so I did, and I noticed the issue could have been caught earlier in the analysis had it been done manually, but... just that initial prompt saved so much time!! Could you create a resource for ensuring AI- catalyzed work has sufficient human oversight for different use-cases? It seems unlikely that AI would hallucinate visualization compared to other code, and relatively easy to check for it, but I'm developing a checklist for future use-cases and don't want to reinvent the wheel...
Thank you for this, I've been getting a little mentorship from a friend who works in Data Science (a more broad sky high level direction), we've been talking about AI and the impact it'll have in Computer Science and Data Interpretation. While it's going to significant and have massive implications for the future, it's important to leverage these tools to expedite the process as for the time being it'll be down to the Analysts/Scientists to interpret and understand the Data as LLMs still struggle with finding more nuance in solutions (their words). Looking forward to expanding my knowledge of these tools and start incorporating them as I progress.
This was a good breakdown. I hadn't used Julius or heard of it before this, so this was helpful.
I am learning analytics for a career change and I love AI for analytics. Its been a great help for when I get stuck or something I am not quite understanding. Or just to bounce ideas off of for why do this or why not do that.