Huy Dao, director of data and machine learning platform at Booking.com, agrees that it’s reasonable to suggest his company is the definition of an insights-driven business. This high-profile status brings a mix of pressures.
“It’s a challenging and at the same time rewarding role because my group and I empower the rest of Booking to make use of the data that we have and then turn it into either insight or actual machine learning and AI capability that powers experience for our users and improves our internal efficiencies,” he says.
Dao joined Booking in August 2023. He formerly held data leadership roles at Zwift and Zwillow. Dao also spent 17 years at tech giant Microsoft, where he latterly held the roles of principal engineering manager and principal data scientist. Now, at Booking, he’s drawing on these experiences to change the business for the better.
“We have many users globally,” he says. “Providing the capability to enable our employees and engineers to develop AI and insights is not easy, but it’s rewarding to see these capabilities rolled out. We have either a direct or indirect impact on some of the use cases that the business is most interested in.”
Embracing opportunities
Dao says the chance to join Booking came at an opportune time. Alongside his family, he was thinking carefully about the new experiences that might appeal. The chance to work in Europe with Booking in Amsterdam was a good fit.
“When they talked to me, I learned about the things the company had already achieved and the potential that data and machine learning could continue to play,” he says. “I thought that focus would be challenging and rewarding. And that’s why my family moved from the US to the other side of the world.”
At Booking, Dao is part of a central group that provides data and machine-learning capabilities to the company. He ensures the data team has the right tools to complete their work. He also aims to make it easy for anyone in the company to build and operate highly governed, high-quality data-enabled products.
“I help them unblock issues that arise,” he says. “I also ensure that we connect with our stakeholders to make sure they are aware of what we have. We work with our internal engineers, and we want to make sure we cater to their needs.”
An engineer by training, Dao’s elevated position means there’s little room in the day job to get involved with coding. However, he still dabbles at home. Dao says professionals who don’t keep up with technology are likely to fall behind – and that’s something he’s eager to avoid.
“When you think about the whole data ecosystem, such as OpenAI and other technology companies, they keep rolling out new things,” he says. “We use quite a few of these services. I want to make sure that I keep up to date with those developments, so I’m not just hearing about the new features, but also playing around with them.”
Integrating information
Dao says Booking has a clear business transformation mission – to make it easier for everyone to experience the world.
“Our strategy is to focus on providing what we call the connected trip,” he says. “We want to make sure that whenever someone takes a trip, we provide a whole, connected experience. So, you book a flight, you book the hotel, you book some attractions – the whole thing should be considered together.”
“We want to make sure that whenever someone takes a trip, we provide a whole, connected experience”
Huy Dao, Booking.com
Creating the connected trip means working across disparate data sets. Much of his team’s effort is dedicated to ensuring high-quality datasets and models are established, utilised and connected effectively. The information they consolidate is used to power important decision-making processes.
Here, Dao points to the Booking Data Exchange, the company’s end-to-end data and machine-learning platform, powered by Snowflake technology. This exchange is used by a community of more than 1,500 data practitioners, serving petabytes of data and hundreds of billions of predictions daily.
Dao says that establishing the firm’s data foundations has been crucial. “One of my biggest achievements has been to come into the company, where we were starting a migration effort from an on-premise Hadoop system into a cloud-based ecosystem,” he says.
“We’ve completed this work in a way that didn’t impact the business. What’s more, the technology adds value by enabling the whole company to leverage the power of this cloud-based ecosystem.”
Exploring innovation
Dao says Snowflake technology plays a key role in his team’s approach. In addition to providing a consolidated and trusted source of information, Booking uses a range of components, including Snowflake Intelligence and Cortex AI, to support its explorations in machine learning, generative AI (GenAI) and agentic AI.
The Snowflake platform forms part of an integrated data stack that draws on other key technologies, including ThoughtSpot for analytics, Astronomer and Airflow for orchestration, Immuta for access control, Arize for machine-learning observability, and AWS for cloud computing services.
“Then we also use GenAI providers, such as OpenAI, Amazon Bedrock and Google Gemini, because we want to ensure that our companies make use of all the technologies and providers available there,” says Dao. “We also use Kafka for real-time queueing and data exchange.”
This integrated approach to technology enables the data team to support the rest of the business as it explores fresh areas of innovation. Whether these explorations lead to new customer services or internal operational efficiencies, Dao says that the aim is to create long-term improvements.
“Booking is known for being one of the companies that pays a lot of attention to experimentation. A lot of things that we do are data-driven because we need to know whether things work or not, and whether we should experiment more using things like A/B testing,” he says.
“I’m very proud to be a part of our approach. We’re producing an ever-increasing amount of high-quality data and are creating end-user capabilities that people love. Internally, our users also leverage similar types of data to ensure that our decision-making process gets better.”
Connecting experiences
Dao says the data platform he’s created allows the company to accelerate the adoption of AI and traditional machine-learning technologies. For example, Snowflake’s Cortex AI provides models and capabilities that mean Booking staff can access data safely and securely for insight-led projects.
“We’ve significantly increased the number of use cases,” he says. “We’re also investing in agentic AI. My group has had a big impact by releasing the first-ever application of agentic AI in Booking, and our CEO mentioned that use case in one of the company’s quarterly earnings calls.”
The system is a partner-to-guest agent that facilitates communication between Booking’s guests and hotel partners, reducing delays and improving response accuracy. The agent automatically suggests responses to each guest’s enquiry. Depending on the message, the agent uses an existing template or generates a tailored answer, saving time and helping hotel partners to reply faster and more accurately.
You need to make sure that you and your senior leaders give your team members what they need and a chance to grow and improve Huy Dao, Booking.com
“And we won’t stop there,” says Dao, reflecting on the data-powered changes that have taken place. “We also have things like our AI Trip Planner and other applications with generative AI. There’s also traditional machine learning for things such as ranking, recommendations, forecasting and some of the pricing components we use. We continually add more of these use cases.”
Dao says his team will focus on how Booking can hone the customer experiences the company provides using the data it collects. Over the next 24 months, he expects further explorations and pioneering developments in AI.
“When it comes to the connected trip, we are on the way, but we are not at the destination yet, when it comes to the experience that we want our customers to have, and that includes leveraging the latest generative and agentic AI,” he says.
“In the future, the customer should expect a great experience for multi-component travel. So, you don’t just book a hotel, you don’t just book a flight, you don’t just book attractions in isolation. If you book several things, we will make sure that the experience is much better.”
Supporting development
Dao’s effective use of technology has reduced operational expenditure costs by 70% through innovative contractor procurement and platform optimisation.
These savings represent a considerable accomplishment, yet Dao says his team, not just the tools they use, is the key to unlocking these successes. “My biggest achievement is being able to continuously work with the team and build them up to deliver the capability that the company requires,” he says.
“When I joined, we had a significant investment in our staff. In the two-and-a-half years I’ve been here, we’ve tripled the team’s size to meet the demands of multiple internal partners within the company. The people component is always the biggest achievement, because without the people, the business goals can be very hard to achieve.”
At a broader level, Dao reflects on changes within the IT industry and says that exploring all the opportunities that emerging technology affords is likely to keep data chiefs on their toes. The seemingly endless roll-out of new models and features means there’s something new to learn every day.
“Data leaders are eager to see how data and AI are used in more elevated roles within the industry and the business,” he says. “This elevation can also make us feel like we’re working in the right area professionally, because I don’t see the trend in other places.”
Yet smart digital leaders of the future won’t just rely on technology. “I think the people leadership component is always there, whether you have five people or a team of 500,” says Dao, reasserting his commitment to staff development. “You need to make sure that you and your senior leaders give your team members what they need and a chance to grow and improve.”
