Airbnb
Airbnb's Price Tips feature helps users figure out what to charge by using machine learning.
Aerosolve is a tool for what the industry calls "machine learning," where software and apps actually get smarter over time.
"We have been operating on the belief that enabling humans to partner with a machine in asymbiotic way exceeds the capabilities of humans or machines alone," writes Airbnb in a blog post.
There's a tremendously growing market for machine learning and artificial intelligence - Google uses it with its new Google Photos to automatically recognize what it's the pictures you upload; Microsoft's Cortana virtual assistant uses the technology to get more personalized to the user over time.
At Airbnb, machine learning works towards a different end.
There are all kinds of things that can affect the price that a user can ask for with their room rental. Is there a big concert or conference in town? Does your apartment have a lot of positive reviews? What neighborhood is it in? Do you have good pictures?
That last one is particularly interesting, Airbnb says. The definition of "good picture" changes depending on who you ask: Professional photographers "tend to prefer pictures ofornate, brightly lit living rooms," but real-life Airbnb customers " seem to prefer warm colors and cozy bedrooms."
These are all things that Airbnb's under-the-hood algorithms have to process just to deliver a suggestion on a nightly room charge.
Aerosolve is what Airbnb came up with to process all that data. And today, it's released under what we call "open source" - a free download so developers can build their own smart recommendation engines based on the work Airbnb has already done with machine learning.
Aerosolve can even automatically figure out which neighborhoods are which on a map based on Airbnb listing data. And it's getting smarter all the time.
Airbnb refers to Aerosolve as "a machine learning package built for humans," designed to be easier to use than other more complicated solutions.