I think it would be possible, why not? No doubt that recent moves in big data will fuel the HR algorithm, and with even more accessible data in the US I think they have all the levers in place. The most difficult hurdle will be addressing the home's "potential" and "condition" considering that the valuation is sight-unseen. 

Considering a market like Toronto where listing a house with an agent can potentially get you a 25-person bidding war, I can`t see something like this working.

However considering some US cities where owners are struggling to sell, this could be a great way for HR to buy into RE assets below market value, and owners to wash their hands.

I imagine that this will have a "guaranteed price" worked into it. So, you can list with them, and if you dont get the offers, then the company will pay the price.

Interesting way to build a long term land holding, using new technologies.

Well played, long term plan, if thats the case.  

I can see this working in areas where houses are under $200,000 and tend to sit on the market for a long time, but don't see it really working in the Toronto market where the average price is now up to $550,000+. Would you really feel comfortable spending $500,000 on a property without knowing that it is actually worth that price (not sure where HomeRun will get their data from but to establish prices they are going to need a lot of data) and any guarantees that there is nothing wrong with the place. I know they mentioned having a home inspector come in, but what about by-law issues, water tests, condo status certificates and all sorts of other areas that need further inspection when buying a home?

When it comes to valuing a house based on data, there is no way to do this. Every property is unique and has various improvements / drawbacks that affect value. Also little things like cleaning a house up before showing it to people can affect what someone is willing to pay for it, which a computer will never be able to tell you.

Will definitely be interested in following this project to see where it goes, but a good realtor should be able to make you the commission back on what your house actually sells for, and also this software definitely won't be free.   

Also, the true measure of value in real estate is what somebody is willing to pay for a product, and in Toronto there are many people that will happily pay much more than something is actually worth. Wouldn't you rather sell your house to those people?

@Matthew Slutsky @Anish Khithani

Doesn't the Zestimate already attempt to assess the value of a home solely on data?

Zestimates have become a lot more accurate since Zillow introduced the feature but Spencer Rascoff has said that he doesn't believe it will ever replace an in-person appraisal.

Details on Rabois' project are scant at the moment but it would be fairly bold of him to think he can outdo the Zestimate in terms of accurately assessing the value of a home using only an algorithm to the degree that it would allow buyers and sellers to completely circumvent agents. 

Am I missing the grander plan here or something?

Sean, from the way I see it, not only will they be estimating the price of the house, but I also assume that they will be guaranteeing that price.

As you mentioned, nothing is really known about this, so it will be interesting to watch.

there is a website that attempts to estimate price of a house for Toronto and surrounding area. http://nestprice.com.

Never go by the details. you must visit the place to get a complete idea of the construction. Attend the events. take a complete study by visiting their site, videos etc. This is one of the best examples: http://www.youtube.com/watch?v=F5nYISmJKrA .
They have demonstrated the introduction of Waterview condominium very well. Every Condo Developer is supposed to follow this idea.

I think that proper investigation is needed before making any deal. You have to see the area, compare the rates and do inquiry about the buyer or seller and dealer as well. As we know that a lot of real estate scams are going on in the market. We must prepare for them to get the benefits and safe deals.

I agree with Zackery, I think you should visit the place personal and hire a experienced home inspection professional. They will analyze the things for you if you can't do this on your own.
http://isaclaims.com/public-adjusters-boynton-beach/  

Data can always be manipulated or presented to represent delusion, reality or absurd unbelievable results, so caveat to any would be users is " be careful the data you choose to rely on and more importantly ask what has been left out of the data collection process aka "filtered out"

It might work with cookie cutter type developments but there are so many intangibles that can sway a potential Buyer one way or the other and influence the ultimate sales price.  The smell, the decor, the neighbours, the buildings orientation on the street, the shape or slope of the lot, feng shui,  what's happening in the news, potential re-zoning etc...  So many factors influence pricing and I always think of it as more of an art than a science.