Dr. Hanna Köpcke: "You Need a Vision to Move the Needle"
An Interview with Dr. Hanna Köpcke, CTO at Webdata Solutions
Perfect pricing policy due to women power: Dr. Hanna Köpcke spun a university project out to found Webdata Solutions in 2012 with Carina Röllig and Sabine Maßmann. Using a special Artificial Intelligence dubbed Black Bee, they convert precise product data available on the internet into information. This market transparency lets online merchants adjust the prices for their products in a more flexible way and stay competitive.
What exactly are the strengths of your Artificial Intelligence?
Maybe you’ve already heard of black bees. These bees are very tough and capable of flying long distances. Now think of the internet as a big, colorful meadow. Our crawlers swarm through the internet like these bees and collect product data like prices, product features and so on. They bring the information back to us and we clean and organize the data to enable our customers to use that data directly.
So, to stick with your image, you’re producing the honey?
The data honey, exactly.
What makes your AI so special to match offers and products?
To perform good market research you need exact comparisons of the prices of one and the same product which is described in various ways on different websites. Not every product has an identification mechanism such as EAN (European Article Number) or ISBN (International Standard Book Number).
Why does your AI make the world a better place?
Our AI offers retailers and brand manufacturers more market transparency. The software has a strong position in the highly competitive e-commerce market, even in comparison with Amazon. Amazon has a whole range of people who only deal with data analytics. They own big databases and therefore, of course, have an advantage. We want to help small and medium-sized companies to stand their ground against the big players.
Might that also be a chance for brick-and-mortar stores?
Sure, but brick and mortars must become more flexible. They must learn to use data in real time to query online prices and to adjust their own pricing accordingly .
Which technology do you use?
Our AI uses algorithms based on machine learning such as support vector machines. It is trained to recognize the most important features and tokens of a product. That way it can differentiate one product from another.
Support vector machines are nothing new. What’s different now?
The methods we use are new. They make for even more efficient learning, so-called Deep Neural Networks. Together with DFKI in Berlin we’re working on recognizing pictures for example. Most of the time, there is only a bit of text to describe a product. Our goal is to use product images to better identify a product.
Can you apply the algorithm to other tasks besides price comparisons?
The technology used to match different images to the same object can be applied to other use cases. You could use it to match job candidates to suitable job openings, which would make it a lot easier to match qualifications and specific expectations on jobs and candidates. Another example: You could query your customer database to guide customers to your website or your call center. Or you can use it to find out if the same customer is listed with more than one address in your database.
You started your company with two other women – that’s very unusual in Germany. What is different in the way women are working together?
In many cases, women are more sensitive to the work atmosphere. To some extent they are better at talking straight when it comes to how people work together. For instance, they might say: "You’re doing really well in this aspect, but you need to improve in another area." In my company, I noticed that men instantly want to know the business case and analyze everything right away. They want to define strengths and weaknesses instead of thinking about a long-term vision. But you need a vision to move the needle.
Concerning new ideas: Is it better if team members are the same gender because all of them think in a simmilar way?
The other two founders and myself have known each other since university. We’re friends, we know our strengths and weaknesses quite well. That's what really matters.
Why do you think having female role models is nessesary in the 21th century?
Young women do need role models to realize that it’s totally normal to occupy top-positions in a company or to lead tech-firms. Maybe an interview like this will inspire more women to study computer sciences and helps them understand that it’s completely normal.
You said that the government already supports future technologies like AI. But do policymakers have enough expertise to understand which projects are important and which are not?
Indeed, sometimes the problem is that policymakers don’t have enough background knowledge to understand a project and to decide if it’s worthy of support or not. But the government has realized that AI is an important technology and wants to make sure that Germany drives the topic forward.
Are today's students eager enough to accomplish what you did? How can we motivate them?
Students are definitely up for working on AI but sometimes they don’t have any practical work experience. Collaboration between companies and universities could help, as can internships where students combine theoretical and practical knowledge early on. That way they’ll understand a lot better what can be done with all the knowledge and will get the "big picture". It also gives them clarity what goals to pursue after they got a degree.
What was the last item you bought on the internet? And did you compare prices?
My cousin’s son was christened on Sunday, and I bought him a gift. I actually did compare prices in advance, but in the end I bought the present in a store nearby.