Artificial Intelligence

Adrian Baetu has been working at Hapag-Lloyd for more than 10 years. After starting out as a System Analyst, he has advanced to become Team Lead Artificial Intelligence (AI). Together with his six-person team in Gdansk, he automates processes at Hapag-Lloyd with the help of artificial intelligence (AI). “AI requires data and a good understanding of the business – both of which we have in great abundance here at Hapag-Lloyd,” Adrian says, adding that, “Things aren’t that way at every company.”

One example of the use of AI at Hapag-Lloyd is the CHOBO recommendation system. “We are familiar with these kinds of tools from online retailers, which provide us with personalised recommendations based on our previous purchases and search history,” Adrian explains. “CHOBO is trained to use our customers’ history to predict where their domestic destination might be and then offer them a good deal on door-to-door transport.”

But AI can do a lot more than that: Bookings and enquiries come to us through various channels – such as by voice over the phone or in text form via email. These completely different types of communication must be translated into a uniform language for our IT systems. The lion’s share of this work is performed by people. “Traditional IT solutions are keyword-based and static, so adapting to changes in wording over time can be very time-consuming and expensive in practice,” Adrian continues. “Using examples drawn from our day-to-day business operations, AI can generate an algorithm that automates complex and ambiguous tasks which would normally require the involvement of a human being. AI solutions can automatically learn from new examples of their human colleagues and then adapt autonomously – thereby creating a so-called learning loop.”

One example of this is the AI solution DAISY, which automates and classifies emails and then forwards them directly to the appropriate employee. “Without DAISY, a team would spend its entire day doing nothing but sorting emails,” Adrian says. “With AI, we are reducing the response time per email from 30 minutes to three seconds.” But this doesn’t mean that AI will replace human workers. “Trivial tasks that are tiring and time-consuming for humans are being automated,” Adrian notes. And AI frees up a lot of human capacity. “Thanks to DAISY, our colleagues can concentrate on the more complex tasks – and therefore increase the speed at which we can respond to our customers.”