Opendoor uses deep learning and crowdsourcing to extract data from home photos
Ian Wong · Modernizing Real Estate with Data Science // Ian Wong, Opendoor (FirstMark's Data Driven) · Feb 3, 2017 · at 6:59
Ian Wong, co-founder of Opendoor, explains how the company converts visual sensory signals from property photographs into structured data for valuation algorithms.
0:00 / 0:11exact quote · 11.0s
▶ Watch the full episode on YouTube → 720p mp4 · rendered on demand · StarZero watermark
“And so practically speaking, what we do is a combination of deep learning, so we have a couple boxes, and Amazon has running CARES and is kind of turning all these photos into structured data, and also a lot of crowdsourcing.”
quote is from the automated transcript, cleaned for reading: filler sounds and stutters are removed, nothing is rephrased. names can be misheard (the analysis reads context, assessments check outside sources). how →
More from Ian Wong
Data science Venn diagrams underplay the importance of software engineering and business
“I think that's the minimum bar to be a sufficient data scientist. A, B it totally underplays the importance of software engineering, and C, that's actually just a small part of what it means to be a data scientist. I think it totally neglects things like busin…”
Opendoor offers home buyers a 30-day satisfaction return guarantee
“With us, we even have something called a thirty-day satisfaction guarantee. So you can actually return the home to us if you don't like it.”
Crowdsourcing is a completely underappreciated technique in data science
“But I'm really hoping that just in 17, as a site know, is going to be the comeback year for crowdsourcing, because we need it. And it's totally underappreciated as a technique.”
Real estate valuation is a wide data problem, not a high-n problem
“And this is not the traditional big data problem. In the traditional big data problem, You have basically a lot of data, right, so a lot of n. Really in this problem we have very wide data.”