Dmitry Alexeenko: “We’d like to empower people to truly immerse themselves in their destination”

September 28, 2017

IT Arena 2017 kicks off tomorrow! The organizers are busy finishing everything, but we keep introducing you to our speakers. This time, we talked to Dmitry Alexeenko, Engineering Manager at Airbnb. Read more about the future of Machine Learning, Airbnb’s plans for expansion as well as Dmitry’s advice for other engineers.


Airbnb tries to build the world’s most trusted community by having a Trust and Safety team working day and night to protect the platform from abuse and fraud. Engineering and data teams have developed a framework that protects the community. How does this system work?

Keeping our community safe, both online and offline, is the most important thing we do. There have been more than 200 million guest arrivals in Airbnb listings to date and negative incidents are extremely rare. Even so, we’re constantly working to improve our platform, our policies, and our protections, because even one incident is one too many.

In fact, Trust and Safety is its own department with offices spanning the globe in San Francisco, Portland, Dublin, and Singapore. Our team is made up of engineers like myself, as well as our 24/7 response agents, data scientists, product managers, designers, law-enforcement liaisons, crisis managers, and victim-advocacy specialists, in addition to policy, privacy, cybersecurity, insurance, and fraud experts – all working together to keep our community safe.

Technology-wise, Airbnb’s risk detection system includes a wide range of Machine Learning models trained on vast amounts of historical data, sets of rules and heuristics built on years of observations and experiments, clustering and anomaly detection systems, and much more.

Online businesses face many risks, with varying exposure. For example, email providers devote significant resources to protecting users from spam, whereas payments companies deal more with credit card chargebacks. What kind of risks does your team face? Were you able to stop all risks so far?

Airbnb is a very complex business with a unique set of challenges, particularly as we seek to bridge the gap between the online and offline worlds.s a result, we face a wide spectrum of risks: from fake inventory to account takeovers, to phishing, spam, chargebacks, and more.On the Trust and Safety team, we work hard every day to strengthen our defenses,  mitigate risk, and prevent bad actors from ever reaching our global community in the first place. Keeping our hosts and guests safe — both online and offline — is the single most important thing we do.

At the end of the day, while we can’t eliminate all the risk in hosting or traveling, we strive to ensure that every host and guest has the best possible experience using Airbnb.

Fortunately, that’s already the case for more than 99 percent of hosts and guests – but there’s always more to be done and we will never stop working to keep our community safe and earn their trust.

What is the future of Machine Learning to you?

I think the power of Machine Learning lies in the data it’s trained on and its applications to real-world problems. I am excited to see Machine Learning becoming more popular, both in the industry and academia, and I’m excited to see how we can continue to apply it as a force for good in the modern world: from improving driving directions, to better language translation tools, to making online marketplaces like Airbnb safer and secure platforms for everyone.

Airbnb works in more than 190 countries and over 65 000 cities. What’s next? What plans does the company have for the future?

We’re now at 4 million listings in over 191 different countries, which makes Airbnb larger than the top five hotel chains combined. Over 200 million people have traveled with Airbnb since our founding in 2008, and on any given night 2 million people are staying at one of our listings around the globe.

Airbnb started as a home sharing community where people can list and book unique accommodations around the world – but where you stay is only one part of your trip. Last November, we launched our Experiences platform, where travelers can truly see a city as a local – rather than a tourist – thanks to the local experts who are sharing their passions with our guests. We’d like to shift people’s travel mindset and empower them to truly immerse themselves in their destination and get an authentic flavor of the city.

Say if someone comes to San Francisco, while they would likely wait around for hours trying to see the Golden Gate Bridge, Fisherman’s Wharf, or eat at Bubba Gump Shrimp – that’s not really a true SF experience. We’d like it to be more about people feeling like locals and experiencing places like Dolores Park, stopping by Borderlands bookstore or going to Tartine Bakery.

How did you become the Engineering Manager at Airbnb? Tell us something about the most valuable experiences in your career. Any advice for other engineers?

I started at Airbnb in early 2014, as a software engineer working on fraud detection. As the company grew we faced more complex technological and business challenges. Systems and processes that worked a couple of years ago didn’t necessarily scale as expected. I have been always interested in providing a vision for folks I work with, helping the team grow, and improving our processes and flows like hiring, onboarding, M&A and working across multiple disciplines like operations, legal and compliance. I think all of that ultimately led me to stepping up into a technical lead role that later resulted in a transition to my current role as engineering manager. I would encourage other engineers to never stop learning, keep challenging themselves, and find an area that they’re passionate about.

Dmitry has been at Airbnb for over 3 years. He is an engineering manager currently leading the payments anti-fraud team whose primary goal is to keep Airbnb’s customers and business safe and secure from the financial perspective. Before Airbnb, Dmitry worked as a software engineer at Microsoft tackling performance problems in Windows and scalability problems in Exchange Server.