robot

Robot Court of Appeal

Let the machine do the thinking for you.




Welcome to the robot court of appeal. Please pardon the mess, I wasn't expecting guests.

This application is a testbed for my experiments analysing publicly available legal information. Initially, I set out to use published cases from the Ontario Court of Appeal to build the corpus, though I may deviate from that depending on how my experiments go. Decisions of the Robot Court of Appeal are generated using the Markov Chain Algorithm, a simple algorithm that I first played with more than 20 years ago. In essence, the RCA renders cases by pressing the middle suggestion button on your text message app -- only the learning set are the decisions of ONCA rather than your sexts. The exact number of cases parsed is set out before you click "Render Decision".

I started this project in Python on the May long weekend in 2020 (during the COVID-19 lockdown) because I missed coding and because I wanted to experiment with AI and data anaylsis and to see if there was useful information that could be gleaned -- in essence, literally "legal hacking". The following weekend, I built this web interface from the simple scripts I put together. Being relatively new to web development, that was a bit more of a challenge, but I think this will serve as a better testbed.

Going forward my plan is to continue to tinker and hack, and I may even update this page if I feel like it.

About Me

I'm a former software and AI developer turned business lawyer in London, Ontario. If you would like to contact me for any reason, you can find me on LinkedIn.