Startup Intraspexion Receives Deep Learning Software Patent – Part 1


editors note: AI Trends met with Nick Brestoff, CEO, Intraspexion, a machine learning start-up that debuted at AI World 2016, where he was showcasing a new deep learning algorithm for the legal industry. The trials and tribulations of a start-up such as Nick’s is an interesting process, and we asked Nick to run thru his experience in obtaining a patent for his product. This is part 1 of a 2 part article. Read part 2 here.

Intraspexion is a Deep Learning software startup and, as anyone can imagine, getting started with a software patent would be a great advantage.

Many would say, Good Luck. Can’t be done. Not after U.S. Supreme Court decision Alice v. CLS Bank in 2014. Don’t you know that Alice made software patent-ineligible?
But on January 24, 2017, the U.S. Patent and Trademark Office issued a software patent to me and it’s rooted in Deep Learning.

So we’ve done the seemingly impossible. But how?

Here’s the back story. I was a litigator for 38 years before I retired in 2014. Before I went to law school at USC in Los Angeles, I received engineering degrees at UCLA and Caltech.
During law school, I met the father of “preventive law,” the late Louis M. Brown. Later, I realized that corporate counsel had access to their own company’s data but nevertheless could not see the risks as they were starting to bubble up, and so could talk about wanting to be proactive, but could not be proactive.

In my book (with Bill Inmon), Preventing Litigation: An Early Warning System, etc. (2015), I showed that the average lawsuit against a business can run about $350,000. Per case. And while I was writing the book, I realized that computer science could now (at last) enable Prof. Brown’s vision.

Having had engineering education before I went to law school, I learned about Deep Learning, and eventually showed that Deep Learning algorithms could get the job done.
I founded Intraspexion, organized a great team of co-founders, and announced Intraspexion’s existence in two public talks (my first) at the AI World Conference & Expo on November 7 & 8, 2016.

And then, on December 12, 2016, the Patent Office sent me an early Christmas present, a notice that it was allowing my patent.

And on January 24, 2016, the USPTO issued my patent for “Using Classified Text and Deep Learning Algorithms to Identify Risk and Provide Early Warning, and gave it a number, 9,552,548. Now you can find it on the USPTO’s website.

Why? That’s Part Two.

by Nick Brestoff, CEO, Intraspexion