How Airbnb, Huawei, And Microsoft Are Using AI and Machine Learning

These three companies are quietly building impressive data-driven products


Machine learning and Artificial Intelligence are two of the most important developments of the past 10 years within businesses. They have been at the core of the success of several companies, from Facebook's advertising policies through to how American Airlines monitors wear of their plane engines.

We wanted to take a look at three companies who are doing some impressive work in the area who are often overlooked for their efforts, either because they are known for other areas or because their competitors sit in the data science limelight.


When you think machine learning, you don't naturally go to 'short term letting'. Renting out rooms and flats doesn't seem like it would be an especially data-driven enterprise, but this couldn't be further from the truth.

Airbnb have not only revolutionized the way that people book their accommodation when in new cities, they have also revolutionized the way the industry utilizes machine learning techniques. A big part of this came from their acquisition of Crashpadder in early 2012, which saw Dan Hill, co-founder of the company becoming product lead at Airbnb and implement Aerosolve, which has since been made open source and available to anybody who wants to use it.

This system creates a pricing algorithm that allows the company to set pricing based on the most popular elements of a property, from the most obvious, like location, through to the more obscure, like the way the photos are taken.

In Dan's own words:

'Here’s where the learning comes in. With knowledge about the success of its tips, our system began adjusting the weights it gives to the different characteristics about a listing—the “signals” it is getting about a particular property.

We started out with some assumptions, such as that geographic location is hugely important but that usually the presence of a hot tub is less so. We’ve retained certain attributes of a listing considered by our previous pricing system, but we’ve added new ones.

Some of the new signals, like 'number of lead days before booking day,' are related to our dynamic pricing capability. We added other signals simply because our analysis of historical data indicated that they matter.

For instance, certain photos are more likely to lead to bookings. The general trend might surprise you—the photos of stylish, brightly lit living rooms that tend to be preferred by professional photographers don’t attract nearly as many potential guests as photos of cozy bedrooms decorated in warm colors.

As time goes on, we expect constant automatic refinements of the weights of these signals to improve our price tips.'

However, it is not only in pricing that the company is utilizing this kind of technology, they are also using it to increase diversity within their teams, detection of fraudulent payments, identify host preferences and even model business impacts of potential product changes. In short, they may well be renting out other people's houses, but they have theirs very much in order.


Often dwarfed by their big competitors, like Apple and Samsung, Huawei have been creating some of the most impressive data science and machine learning facilities in the world called the Noah's Ark Laboratory, concentrating on:

- App Recommendation & Search Engines

- Deep Learning for Natural Language Processing

- Intelligent Banking Solutions

- Intelligent Help

- Learning to Match

- Mining from Big Graph Data

- Natural Language Dialogue

- Stream Data Mining

- Telco Big Data

- Educated Artificial Intelligence

The lab they have created has also had some big wins, not only for them, but also for the networks on which their phones run. For instance, their work has helped carriers across the world to reduce their pay-as-you-go customer churn rate from 10% to 6%. They have been working on a machine learning driven network control system, which will achieve automated network traffic control. According to 'Tests indicate that Network Mind is up to 500% more efficient in realizing KPIs such as task completion or policy generation compared to existing template or heuristic algorithm-based optimization methods. Network Mind is also over 50 times more efficient when analyzing paths of large optical networks, which has the potential to reduce the time it takes to analyze use cases such as optical network failure prevention from 5 hours to as little as 6 minutes.'

So whilst their competitors are certainly doing some exciting work, Huawei's work is going to have a real and direct impact not only on their bottom line, but also on the wider technology and telecoms community.


It is not so much that Microsoft are being ignored in their efforts, but when it comes to their machine learning and AI initiatives they have often been seen as trying to play catch-up or making mistakes. Their most famous AI driven embarrassment coming from their AI bot on Twitter, Tay, who became racist and bigoted within 24 hours of inception, denying the holocaust and inciting race wars within 12 hours.

However, they are also doing some amazing work and perhaps the incident with Tay, which learnt from what was said to it, is more an indictment on humans than it is on Microsoft's machine learning skills. We have seen with their roll out of Azure that they have some strong pedigree in the area, with the system constantly improving and getting closer to what's currently offered by IBM Watson, which has set the current benchmark of AI and Machine Learning systems.

Azure is not only having an impact in the US, but across the world, with Andhra Pradesh, a state in India, embracing the platform to improve themselves. One of the first successes that the partnership had was to predict the dropout rate for schools in the state using machine learning, with over 10,000 schools taking part in the experiment.

They have also worked with doctors from the LV Prasad Eye Institute (LVPEI) where Microsoft has utilized medical data from millions of treatments across the locations where the LVPEI operate. This data was then used to help predict the chances of success and improvements from surgeries, allowing doctors to recommend both the treatment that would help the most or even that would provide the biggest improvement for the least money. 

Bean small

Read next:

City of Chicago: An Analytics-Driven City