Written by Andrew Seale
While studying for his double degree at University of Waterloo and Laurier University, Ron Glozman would’ve given anything to split himself in two so he could tackle both course loads at once. Instead, the computer science and business major came up with a more realistic solution – he taught a computer to turn his thousand-page textbooks into one-page chapter summaries.
“I did the math and I'm like, look a textbook is at least 1,000 pages and the final exam is maybe 10 or 20… so (that’s) one percent, what do I do with the other 99 per cent?” says the founder of Chisel AI, which provides AI-based solutions for the insurance industry. “I ended up building that solution over a period of four and a half months.”
At first, he kept it for himself like some secret superpower. But it raised eyebrows. His friends in class started prying, they wanted to know how he was acing his courses despite not coming to class.
“I told them I have this app where I can take any textbook, tell it what chapters are on the midterm or on the final exam, and it's just going to give me a summary of those chapters,” he says. The secret was out, they wanted access. So he threw it on the app store. The first day it got 185 downloads. “Every friend or acquaintance I knew was like, let me download this.”
Within two weeks it had made it to the front page of Product Hunt, a website where users share new products. That day alone several thousand users signed up from 44 of the top universities in the world including Princeton, Harvard and Yale and 33 countries from India to Brazil. “Literally on every continent… clearly, I found success,” says Glozman. “Later that year it was called one of the best 50 apps for students of all time.”
But it proved impossible to turn a profit. Students were cash strapped and didn’t want to pay. He tried selling it to the university but the procurement process moved at a glacial pace.
RBC was holding the first ever machine learning conference and invited Glozman to talk on a panel about AI. After the talk he received an email from a senior VP at one of the biggest insurance companies in the world saying “I just saw you give a talk, if you have five minutes, I think this could really help the insurance world.”
Glozman wasn’t that well-versed in the sphere or even the company. “I did a quick Google of the company name,” he admits. But he quickly realized the legitimacy of the situation. He went in to meet the VP and she asked if he could do some consulting and how much it would cost to build a similar solution for the insurance industry.
“I tell her the biggest number that comes into my head and she just signs the check and I was like, oh my god, I went too low… that was my first lesson: if someone doesn't negotiate with you, you underpriced it,” says Glozman. The years since have been a whirlwind.
Glozman’s Chisel AI has evolved into a machine learning leader in the insurance sphere, designing software capable of reading insurance forms in milliseconds and extracting key policy information for the world’s largest brokers.
He’s participated in Aviva Insurance’s accelerator at the DMZ as well as graduated from both the Creative Destruction Lab at the Rotman School of Management and the Next 36. The company also raised one of the largest seed rounds in Canadian history.
Glozman says Toronto provides the perfect base to siphon the top talent from two of the world’s leading AI departments – the University of Waterloo and McGill University in Montreal.
“Arguably two of three forefathers of modern day deep learning and machine learning are Canadian… it was founded here and is taught very well at those universities,” he says. But it’s that same renown that keeps Glozman up at night. “My biggest fear is brain drain, we train all these really smart and capable developers, product managers, whatever, lots of skills and then they end up leaving Canada,” he says.
He’s made stopping that brain drain his mission, working alongside other startups and policymakers to try and entice people it stay in Toronto. “It has to be a mutual effort with the government, the startups, universities, and corporations working here to make it extremely sweet for people to stay,” he says. “I don't have an answer yet.”
It feels lightyears away from the university days when his AI-driven textbook skimmer was simply a superpower he’d yet to unleash into the world. “If you asked me when I started six years ago,” he says. “This is probably not where I expected to end up.”
Photo credit: Cameron Bartlett (www.snappedbycam.com)