Our people is on recommendations from user specific insights based on historical legal cases. Let me 
explain what we propose. While fighting a case in the court of law, the parties contesting the case or the 
lawyers need to be adequately prepared and this can take a lot of time and effort. Take divorce cases for 
example. These are painful and emotional for the parties involved if they want to look at the past. Records 
to understand how cases have preceded the legal parlance may be difficult to grasp. As for the lawyers, 
they may have to go through dozens of historical cases for the case at hand. These historical records may 
be two pages or 20. Reading and grasping may take days, and concentrated efforts will be required to sift 
the main points from the journal verbiage. There are search tools providing general guidance regarding any 
situation, but they do not cater to each user profile. Every situation may be different. And may require 
different actions from the parties or lawyers. Our research team saw the need for personalizing and 
contextualizing and pulling up a set of recommendations that can guide a parties or lawyers preparations 
and also show the case trends to the plaintiff and defendant. The solution we propose asks the user to 
answer 4 questions in a multiple choice mode. It curates the provided data and generates statistical 
insights. Statistics gives a feel of how many cases in the past with similar elements have proceeded to a 
particular outcome. These statistics fill up the blanks. In the predefined templatized recommendation, 
we model the concepts and categorizations of the appeal facts and the reasons of the decisions using 
modeling, natural language processing and machine learning techniques to let the solution pull up 
recommendations. We feel that there has been hardly any attempt to personalize recommendations to this 
level, especially in the legal realm. We are sure this approach will be useful to lawyers, counselors and 
even the general public to be more aware of legal positions.