Julius AI

Today we’re excited to announce Persistent Files, our most requested feature. Maintain your flow and preserve your workspace with our new week-long storage. Now you can analyze data, step away, and return days later to pick up right where you left off.

We're excited to roll out Custom Containers, a powerful new way to tailor your Julius workspace. With options for extended sessions and increased memory, Custom Containers lets you tackle complex, data-intensive tasks with ease.

"Some new data analysis tools such as (the impressive) Julius AI can run complex and sophisticated statistical analysis, thus enabling advanced statistical analysis without specialized training." We appreciate the mention Avi Staiman!

New update: convert a thread into a Workflow with a click. Easily turn any of your Julius chat threads into a templated Workflow to reuse or share with others.

New in Julius: Workflow Creation. Have a periodic report or recurring data analysis task? Offload it to Julius and create a repeatable, templated AI workflow.

We're excited to announce Workflows in Julius. You can now run repeatable, multi-step analysis on your data using templated AI Workflows.

I use Julius AI to analyze and visualize large data sets when doing research for CRO clients at DTC Pages. The most powerful and important prompt I use before running any analysis is the following: "Find data inconsistencies and improve data quality. - Find data inconsistencies - Improve data quality. - Generate a new CSV file with the data corrected." Thank me later, but this will help you're not working with trash data.

A constant complaint from law firm clients is that M&A due diligence reports are too long. However, there are many obstacles for firms to overcome this. For example, it is extremely difficult for firms to process the volume of information and data (often provided by the seller as a ‘data dump’) in a concise report using the current method of manual reporting in a Word document. Furthermore, reports are usually prepared under extreme time pressure and budget constraints. Plus, it’s always hardest to present something complex, in a simple, understandable way so often this is overcome by including more detail, rather than less. When I was at Gilbert + Tobin, one of the ways we tried to make the reports shorter and easier to digest was by adding visualizations to break up the text. We would visualize the insights uncovered from the document review. They weren't ground breaking - mostly bar or pie charts and tables - but it made the information easier to read and we received good feedback from clients on this approach. However, one of the issues we still encountered was presenting the visualizations so that the information contained within it was still understood by the reader. Working with Word reports, it has significant limitations and visualizations have to be included as static images, rather than being interactive. Another example was when we would use a table to summarize the top customer contracts and each column was used to summarize a key term. Sometimes there were so many columns in a Word table that each column became so narrow that you could hardly read the content. To try to make reports shorter and more interactive, I have been playing around with Julius AI to visualize contract data. I uploaded a spreadsheet of metadata from key clauses (including termination and change of control) of 100 material contracts reviewed by 10 document reviewers using Due. I asked Julius AI certain questions about the data and asked it to visualize the results. It converts the data into visualizations within seconds that can be added to digital reports. The key step here was to have a checklist of all the data points to be extracted from the documents so all the reviewers were extracting the key information consistently. It might be possible to do this analysis in Excel, however Julius AI also gave analysis on the results (for example, it understood that the joint venture decision-making threshold was about which joint venturer could pass decisions on its own and/or veto decisions). This is exactly the sort of analysis that clients are wanting when considering an acquisition of a joint venture interest. I have prepared a short video using Arcade showing a few examples of the visualizations I prepared that can be added to a digital report, instead of long passages of text. Feel free to reach out if you have any other examples of data that can be visualized in a DD report (I am sure there are lots more!)

Claude 3.5 Sonnet is now live in Julius :) #claude #anthropic #ai #artificialintelligence

Introducing a major upgrade to our charting & graphing capabilities: Interactive Charts. With only a prompt, you can now generate dynamic, presentation-worthy charts with improved styling and easy, one-click editing in Julius.