At Infare we collect 2 billion data points every day, and it’s the job of our data scientists to conduct creative inquiry and analysis. In other words, to crunch the data!
We took time recently to pick the brains of one of our top data scientists, Vadim Skritskiy, to bring you this Infare Insider story about ‘a day in data scientist life’, the latest developments of data science in the travel industry and how Infare’s data science team is mastering that.
Can you describe a ‘normal’ day at Infare for data scientists?
It’s very interesting and challenging at the same time, delving into maths and formulas, looking at charts and graphs, challenging the status quo, and lots and lots of thinking…
A big part of the work is building a strong data foundation from multiple data sources, including machine learning and artificial intelligence.
Having airline business knowledge is key to data science projects at Infare. That is why we regularly interact with Account Managers so that we fully understand the needs and demands of our airline customers, as they have a strong business perspective, extensive revenue management and airline competition knowledge.
Early into our projects, we also involve selected customers to gather essential feedback. This guides us in the right direction from an idea to building valuable best market-fit products.
My North Star is creating true innovation. Not speculating on products or services but producing value-adding solutions.
Vadim, tell us about your career at Infare
In late 2017, Infare opened its 3rd office, the Development Centre in Krakow, Poland. It was a true honour to be the first to join at this location and to take a part in building up the office from the ground up. Shortly, in just a month we were two and in the following months the Data Science team was fully formed. Now, we are over 20 people in the Krakow office!
The beginning at Infare was full of learning for me, both professionally and personally. My background is in Applied Mathematics with many years spent working on various Machine Learning and AI projects; from language processing to building recommendation engines. However, I have never worked in the aviation industry before. Joining Infare allowed me to explore the whole new spectrum of possibilities this data-rich aviation industry has to offer for a data scientist like myself.
What is your take on data science in the travel industry?
Digitisation has come a very long way in the airline industry. In 1960, American Airlines’ Sabre flight reservation system was processing 84,000 telephone calls per day and storing 807 Megabytes of reservations, flight schedules and seat inventory. It’s grown just a bit since then!
Today the aviation industry is one of the most data-rich industries on the planet. The landscape of how people search and buy flight tickets alone has changed dramatically. Since Infare was founded almost two decades ago, our airfare database has grown to over 1 trillion historical airline ticket prices and it is expanding by 2 billion airfares a day.
As a data scientist, it excites me to see the huge potential and value this data source has to offer to the airline industry. Our Data Science team spends a lot of time exploring the true value of this data and what products can be built to maximise its value.
What excites you about the future of airline competition from the data science perspective?
Airline competition is very complex as of course it is linked to other businesses. For example, data on nearby airports, code-sharing agreements and geographical data can often be undervalued. Then of course there is data on population trends, travel time to airports, purchasing power, ground transport prices, and passenger flow on stop-over routes. They (and many other factors) all play a part in the airline marketplace. There are plenty of options available to travellers on ‘getting from A to B’. Enabling competition transparency is vital and Infare is helping airlines achieve this.
Another aspect of data science in the travel industry I find fascinating is our ability to detect airline industry anomalies at the right time. In 2017, Vilnius airport in Lithuania was under construction. Some airlines decided to temporarily stop operating their routes from that airport. However, one proactive and data-enabled airline quickly moved their operations to the smaller airport of Kaunas. They increased capacity and earned substantial revenues and profits in just under 6 weeks by offering their customers an alternative. It is just one example of how proactive airlines can use data insights to create business opportunities and profit. It’s great to see our customers benefit from our products and our daily work as data scientists.
Vadim was also a panellist representing Infare at IATA Aviation Data Symposium 2019 alongside representatives from Revenue Management Group Inc., Lufthansa Group and Skyscanner. The panellists addressed the topic of “Matching supply and demand – data driven network optimisation” and discussed how airlines can better react to changes in demand and profitability, airline industry anomalies and airline competition through new techniques and real-time access to data.
Talking about competition, can you tell us more about the competition you won at the IATA Aviation Data Symposium?
IATA Aviation Datathon is a competition within IATA Aviation Data Symposium where data scientists not only debate but take action on site and propose solutions. It is a one day workshop-like competition leveraging data science in the travel industry capabilities in areas where data is already available and accessible.
I competed with other industry experts who wished to address social media sentiment analytics in travel. I believe there is a huge benefit to addressing travellers’ sentiment with actual airline pricing data. As an outcome, I built a social media sentiment model for airlines and airports that decodes tweets and emojis. Winning this competition only confirmed my hypothesis that there is the untapped potential of combining different data sources like social media and airfare data. One use of this model is that an airline or an airport employee can quickly aggregate and analyze hundreds and thousands of tweets to identify a general customer sentiment; an aspiration to travel, attitude towards prices etc. Not only is it my honour to bring this award to Infare but I am also excited to see this model published in IATA’s data science whitepaper, 2019 Q3 version! This is the start of a new conversation for the travel industry and potentially new innovation on the way.
Feel free to reach out to Vadim directly at email@example.com to discuss the latest Infare news.
YOU MIGHT ALSO LIKE:
When we think of pricing intelligence, we tend to see Amazon as the trail blazers, with algorithms that reportedly change […]Read More ›
Get a fresh perspective from an Infare Insider. This month, we speak to Infare’s User Experience Designer Elina Talashka about […]Read More ›