He looked at a $30 trillion housing market still running on paper and incomplete records, and decided to teach a machine to value every home in the country.
The name is the tell. A canary is an early-warning system, the small thing that notices danger before the rest of the room does. Jeremy Sicklick named his company HouseCanary because he wanted to build exactly that for the biggest, slowest, most paper-bound market in America: residential real estate.
Today he runs a San Francisco company whose machine-learning engine values and forecasts more than 136 million homes - close to the entire U.S. housing stock. The pitch is deceptively simple. A traditional home appraisal takes two to three weeks and one human with a clipboard. HouseCanary aims to deliver a trusted valuation in seconds, and to be more accurate doing it.
"We've focused on automating up to 70% of home valuations to be more accurate and faster than human appraisals."
Sicklick did not come from real estate, and that is the point. He started his career in the 1990s as an accountant and consultant at Arthur Andersen, then spent years as a Principal at Marakon Associates. By 2008 he was a Partner and Managing Director at the Boston Consulting Group - and that is where the origin story actually begins.
The housing market was collapsing. Sicklick, trying to make sense of billions of dollars in property values for clients, ran into a wall that had nothing to do with economics and everything to do with plumbing. The records were paper-based. They were incomplete. There were no common data standards for the single largest asset class in the country. The market that had just nearly broken the global financial system was, underneath the hood, a filing cabinet.
So in 2013 he left the partner track to fix the plumbing. He found his co-founder in Christopher Stroud, then a 27-year-old doctoral student in statistics, who became HouseCanary's Chief of Research. The two of them set out to digitize and index information about every home, every mortgage, and every neighborhood in the United States.
"To give everyone more information about more homes so they can buy and sell homes more easily."
The thesis underneath all of it is borrowed from a market that already works: equities. The stock market is liquid and transparent because everyone can see a price. Housing is neither, because almost no one can. Sicklick's wager is that if you can value a home instantly and trust the number, you make the whole financial system safer, more transparent, and more liquid - and you lower the risk baked into the largest pile of debt in the country.
It is a consultant's instinct applied at national scale: find the part of the system everyone tolerates because they assume it cannot change, and change it. The clipboard, in this story, is the thing that cannot survive.
Starts out as an accountant and consultant. The numbers habit forms early.
Principal at the strategy firm, advising on value creation.
Partner and Managing Director. Stares into a collapsing housing market and finds it runs on paper.
Co-founds the company with statistician Christopher Stroud. Begins indexing every home in America.
A jury finds a Quicken Loans affiliate misappropriated HouseCanary's trade secrets, awarding $706.2M.
Named a HousingWire Vanguard and Inman Power Player; launches CanaryAI Beta, a chatbot for real estate data.
In 2018, a jury awarded HouseCanary $706.2 million after finding that Amrock - a company in the Quicken Loans family, formerly Title Source - had misappropriated HouseCanary's valuation trade secrets, systems, and data. A judge later pushed the figure toward $740 million with interest and fees.
The case turned a quiet data company into headline news, and made a blunt point: the property intelligence Sicklick spent years building was valuable enough that a much larger company allegedly tried to take it.
Note: the verdict was contested and subject to extended post-trial appeals.
Modern machine learning gets to a more accurate valuation than a human appraiser - and a lot faster.
Reliably automating home valuations replaces a 2-3 week appraisal with instant, trusted valuation at a fraction of the price.
Automated home valuations help make the financial system safer, more transparent, and more liquid.
Make real estate data analytics easily accessible and digestible for everyone.
"Canary" is no accident - it is the bird in the coal mine, an early warning for the housing market.
Before AVMs and AI, he was an accountant at Arthur Andersen in the 1990s.
His partner Christopher Stroud was a 27-year-old statistics PhD student when they started.
Lives in the Bay Area with his wife, three children, and one notably energetic dog.