This year marks two major events: We are entering into the Data/AI industrialisation decade and Data Innovation Summit celebrates the 5-year mark journey. Two significant reasons why we are extremely excited to announce the Celebrate edition of the Data Innovation Summit.
But we wouldn’t have been able to come this far without our speakers who have selflessly shared their stories, experiences, knowledge, challenges and successes with us and our delegates. We’ve come together this far together and we should celebrate together.
For this purpose, we reached out to our past speakers from previous editions of the Data Innovation Summit and asked them to share with us their journeys with data and analytics, their views on how the data and AI scene has transformed during these 5 years.
The speaker interviews will be presented in a series of interviews with past DIS speakers covering the topic “Then and Now, and Outlook for 2030” – Starting with themselves and their journey in the past 5 years; where were Data, Analytics and AI 5 years ago, what was discussed then, where they are now and what can we expect in the next decade to come.
The first one to break the ice is Saulius Valatka, Technical Lead at Adform.
Hyperight: Hi Saulius, it’s great to catch up again! You were one of the speakers at the 1st edition of Data Innovation Summit 2016. Just to refresh our memories, please tell us a bit about yourself and the company you are coming from.
Saulius Valatka: Hello and thank you for catching up! Back in 2016, I had recently re-joined Adform, one of the biggest ad tech companies in Europe, where I still work to this day. My area of focus has largely remained the same, namely, I work on building data processing pipelines with the goal of making Adform an increasingly data-driven company. While my team and I work on reliably and efficiently delivering the data, our data scientists and analysts make use of it to build models for bidding in real-time ad auctions, detecting fraud, forecasting, segmenting audiences and other areas.
Hyperight: Next year we are celebrating our 5th anniversary. A lot has changed with data and advanced analytics during these 5 years. From your point of view, where do we see the biggest changes and advancements with AI and data we have had?
Saulius Valatka: Indeed, the first Data Innovation Summit took place in March of 2016, which I remember being a very optimistic time for AI, as just days ago DeepMind’s AlphaGo had defeated a human professional Go player in a five-game series. This was a major stepping stone for AI, the future looked very bright and in the years that followed, we did more applications of AI in other areas like self-driving cars, diagnostic medicine and natural language processing. Unfortunately, it wasn’t all sunshine and roses, we also had to face the hard truth that these are powerful tools that have to be used very carefully, as we saw with the Cambridge Analytica case, numerous data breaches, surveillance scandals and the increasing public concerns and awareness regarding data privacy.
Hyperight: At Data Innovation Summit 2016, you presented the topic “The Secret Weapon Of Digital Advertisers”. As you are coming from a digital advertising background, can you give a short recap of how data and analytics used in advertising in 2016 and how they are used now?
Saulius Valatka: Back then we had just a few products using data at their core, the real-time bidding engine being the prime example, which I presented in some detail at the summit. In short, the goal of real-time bidding is to spend an advertisers budget optimally by picking the most relevant ads for users in real-time. Obviously, the sophistication and efficiency of algorithms making these decisions have increased dramatically over the years. We have also started applying data and machine learning in other products that we build by enriching their functionality with even more advanced analytics, forecasting and other “smart” features.
The hardest thing in working with data is that it’s always growing, so even with faster hardware and better technology, it’s always a very close race to keep up.
Hyperight: Data was king then, and it still holds the throne. Now let’s talk more specifically about data science in advertising. What has changed in terms of the data science tools and technology used in advertising?
Saulius Valatka: Technology-wise I believe a lot of focus in the last few years has been on speed, for example, stream processing has almost completely replaced batch processing, so today we can react to data faster and adjust our models in real-time. Querying and analyzing data has become easier, we now have more database technologies to choose from, such as Vertica, ClickHouse, Druid, Presto, all of which have made tremendous leaps and bounds during the last few years. Advances in GPU technology have allowed us to train bigger and more powerful models in shorter amounts of time.
Hyperight: What about the challenges? Does more advanced technology comes with even bigger challenges or is it easier now to work with data in advertising?
Saulius Valatka: The hardest thing in working with data is that it’s always growing, so even with faster hardware and better technology, it’s always a very close race to keep up. Overall, I think the industry matured a lot in the last five years, we learned a lot, solved a lot of hard problems, but there are always new challenges waiting around the corner. For example, an unexpected complication in working with data during the last few years has been the General Data Protection Regulation (GDPR), which basically means that we have to be able to retrieve and clear all data belonging to a particular user upon his or her request. While this is obviously great for users, it turns out to be rather difficult to implement technically. One thing is for sure though: working with data is still as fun and exciting as five years ago.
Hyperight: Talking about the decade to come, what are your future outlooks regarding data and analytics in advertising in 2030?
Saulius Valatka: Obviously, the importance of data will only grow in the future, which at the same time means opportunity, but also entails responsibility. With the current pace of advances in AI, it is very hard to imagine what will be possible in a year, let alone ten years. Either way, I am very excited to see what the future holds in store for us.