Julius is a great AI tool for analyzing and visualizing datasets

This article is republished with permission from Wonder Tools, a newsletter that helps you discover the most useful sites and apps. Subscribe here.

Julius is a promising new AI service for analyzing data. I’ve been using it to make sense of—and visualize—my analytics dashboard and find meaning in giant datasets.

You can use it for virtually any type of business or scientific data, or simply to categorize survey responses or interpret spreadsheets. Read on for what’s notable about Julius and other new approaches to AI data analysis.

How I’m using Julius

Julius has helped me make revenue projections and identify reader retention patterns. It can turn anonymized analytics data into specific actionable observations. Substack, Google Analytics, and other platforms offer limited dashboards for understanding readership trends. I have neither the budget to hire a data analyst nor the time to do deep analysis of large datasets without help. So Julius is handy at providing me with insights I can further explore. I think of it as generating idea leads—or data idea seeds.

It’s remarkably versatile. You can use it to break down a scientific or geographic study, financial data, or anything else. It runs Python code, so you can replicate its analysis. It builds on the top models from OpenAI and Anthropic, tuning AI specifically to focus on data.   

What you can use it for

Here are examples Julius’s team shared with me, now that there are more than a million people using it: 

• Academic research data: Biology experiment results, bioinformatics datasets, psychology doctoral thesis data, and survey data analysis.
• Business data: Google Ad reports, cybersecurity data, a product manager’s product usage and behavioral data, and sales forecasting.
• Data science analysis: Predicting housing prices based on economics data, clustering customer segments, and detecting credit card fraud. 

How Julius differs from ChatGPT

• Julius lets you upload huge data files—up to 8 gigabytes on paid accounts.
• You can use Python as well as R, a preferred coding language for academic researchers.
• With Julius, you can access extra shared server computing horsepower—both CPU + RAM. Basically, it can crunch more numbers faster than other AI chat services, if you’re on a paid account.
• You can repeatedly reference data files once you’ve uploaded them. If you’re using Julius for free, data may not be stored, though. I’ve had to re-upload files.
• You can install special analysis packages/libraries for advanced projects. 

Here are some other AI tools for data analysis:

Wobby

Wobby lets you import your own data, paste in a URL, or search for data sources. You can upload a .csv or Excel file or even copy and paste something from a spreadsheet. You’ll soon also be able to upload a PDF.

Once your data is uploaded you can chat with it, visualize it, or draft a document summarizing key data insights with AI assistance. 

The interface centers around the creation of documents. You can create standard documents or landscape-style slides. I prefer creating docs and slides in other tools, and I prefer the Julius interface for exploring datasets.

Wobby feels a bit rough around the edges so far in my testing. It has a lot of potential as a way to turn raw data into reports you can share with colleagues. I appreciate its versatility in letting you choose to upload or search for data. If you’re frequently generating internal reports about company data, it may save you time.

Cofounder and CEO Nathan Tetroashvili recently told me about news organizations using Wobby to streamline data analysis in Europe. In one case, a small news team used Wobby to visualize local election data. They created custom data visualizations for each locality that would have been much more time consuming and difficult to produce with their prior tools.

• Pricing: After a 10-day free trial it’s $36/month billed annually for individuals; more for teams.
• Privacy: Wobby is incorporated under Belgian law and complies with GDPR rules that help safeguard data privacy. Read its policy for further details.
• ">Watch a helpful, 8-minute  ">demo video showing Tetroashvili using Wobby to analyze, visualize and draft a report about the gender breakdown of national parliaments in Scandinavia.

Bigdata

Bigdata is useful for analyzing market info, pricing data, job analytics—who is hiring—and other financial sources and stats. It’s not for the casual user. Plans run $50 to $100/month billed annually, so I won’t be maintaining a subscription. If you’re in finance you can likely expense it. If not, you can try it free for a month—like me—to see how AI analysis is progressing and to marvel at the kinds of questions it can answer.

ChatGPT

ChatGPT works surprisingly well for analyzing big datasets. Tasked with breaking down global energy consumption data, it quickly produced a detailed analysis and a series of informative visualizations. It’s also great for making sense of survey data.

Be sure to strip out names, email addresses, or other identifying or personal info before uploading data for analysis. On the free plan you’ll have limited access to data analysis. For particularly tricky datasets, try a Custom GPT called Data Analyst.

Claude

Claude has also been helpful for data analysis. I use Claude Projects every day—here’s why I find them useful—so I appreciate being able to analyze data within the context of a project Claude already understands.

• Limitation: Unlike Julius and ChatGPT, it hasn’t accepted large data files I’ve tried to upload. But it has provided useful insights when I’ve given it survey data and other smaller datasets.

Perplexity

Perplexity is known as an AI-enhanced search engine, but it also functions well for analysis of uploaded data files. Here’s how and why I find Perplexity so useful. Now I’m using it for data analysis as well.

• How it works: Before typing in a query, attach a data file for Perplexity to focus on. When I gave it a dataset of board game reviews I found on Kaggle—Perplexity suggested numerous useful avenues for research. It then provided clear, easy-to-understand analysis. Here’s the thread and a screen recording.

This article is republished with permission from Wonder Tools, a newsletter that helps you discover the most useful sites and apps. Subscribe here.

https://www.fastcompany.com/91235755/julius-is-a-great-ai-tool-for-analyzing-and-visualizing-datasets?partner=rss&utm_source=rss&utm_medium=feed&utm_campaign=rss+fastcompany&utm_content=rss