Skip Navigation

Data Science
at Deutsche Post DHL Group

The logistics industry and data science were made for each other - and, at Deutsche Post DHL Group, we are serious about it.

Join our journey

We employ over half a million people around the globe across postal services, international express delivery, forwarding and contract logistics. Not surprisingly, our activities generate a lot of data. We believe that our data is the key to unlocking untapped potential in our company – and we are looking for ambitious data scientists and data engineers that can help us enable our digital transformation journey.

Machine learning, mathematical optimization and data engineering are our domains. If you would like to know more about us, check out the information below – and don’t hesitate to give us a call. We are always happy to answer any questions you may have.

What we do

  • Predictive Analytics
  • Prescriptive Analytics
  • Routing Optimization

Predictive Analytics

We love employing machine learning to give our business a glimpse of the future. What is the warehouse demand next month? At what time will shipments be delivered? Is a particular customer likely to change to a competitor? The questions we get from our business are challenging – and their requirements are ever-changing. The variation in the problems that we solve for our business is also one of the key reasons why our data scientists must be comfortable with time-series forecasting, classification and statistical analysis.

Forecasting Letter and Parcel Volumes

As the national postal service for Germany, last year, we delivered 4.6 million parcels and 59 million letters per working day. Accurately forecasting these volumes is difficult but also crucial for our business. Previously, our letter and parcel business was forecasting their volumes based on human input only. Because of the expertise of our colleagues in the business, their forecasts were generally quite accurate; however, by employing machine learning, and combining this with the knowledge of our business colleagues, we made significant improvements to their forecasts.

Prescriptive Analytics

Prescriptive analytics is the natural successor to predictive analytics. When we perform predictive analytics, it is (not surprisingly) to tell something about the future. After we have generated insights from our predictive models, we typically venture into prescriptive analytics to help our business partners make the most of this information.

Staff Scheduling

In the earlier predictive analytics example, we first helped the business to forecast volumes using machine learning. With this information, we worked closely with the business to create a prescriptive staff scheduling algorithm to optimize the working schedules for our colleagues in the distribution centers. Thus, the volume forecast helped to meet several objectives for our business.

Routing Optimization

As the biggest logistics company in the world, we can also boast one of the most extensive transportation network in the world. However, size comes at the expense of complexity and our data scientists play a significant role in ensuring that shipments are transported both timely and efficiently. Delivering logistics services requires much more information than just point A for pick-up and point B for delivery. What is the most efficient route? Do we optimize on time or cost? What mode of transport should we use? From where should we fly? How do we consolidate? These are just a fraction of the questions we are expected to solve.

Optimizing deliveries

Think back to our previous examples in predictive and prescriptive analytics. We have forecasted the incoming volume of letters and parcels; we have optimized the staff working on-site, now it’s time to deliver. While our previous exercises generated significant value to the business – the delivery of letters and parcels is of utmost importance to any logistics company. The natural next step was therefore for our operations research colleagues to optimize the routing of our delivery personnel to ensure that the whole process is as efficient as possible. The Traveling salesperson problem is omnipresent in logistics and we are always looking for talented operations researchers to optimize our routing.

How we do it

The DPDHL Data Science Organization

The data analytics teams at DPDHL are organized dynamically. In corporate, the Center of Excellence is responsible for divisional and cross-divisional end-to-end projects, training, recruitment and information exchange. In addition to the Center of Excellence, our business units have their own Data Scientists focusing on their specific challenges. An exchange of ideas and manpower between the Center of Excellence and the business units is common and highly encouraged. In fact, we often see projects that originate in the Center of Excellence move to the business units and vice versa. What is important to us is that we create the best analytics solutions for DPDHL.

To ensure that our projects, regardless of their origin, can be scaled from pilot to production, we have a close working relationship between our data scientists and our data engineers.

Corporate Data Lake

Data is an asset – and, in DPDHL Group, we consider data to be as important an asset as our hubs, planes and trucks. It is imperative for us that all data scientists in the Group have access to data from a reliable and easy to reach source. Our corporate big data infrastructure, more colloquially known as our Data Lake, is currently under implementation – so if you are a data engineer looking for an exceptional opportunity, this is it. We aim to enable our data scientists to deploy their big data and machine learning solutions, with a single click, without worrying about scalability.

Our Tools

At DPDHL, we evaluate each of our projects individually and select the best set of tools that can help us solve the business problems we are facing. These tools can be varying from small open source libraries to large commercial products. Some of our most frequently used tools can be found below:

  • Programming Languages
    • Python
    • R
    • C++
    • Java
    • Scala
  • Machine learning
    • Scikit-Learn
    • Tensorflow
    • Keras
  • Operations Research
    • Cplex
    • AIMMS
    • LocalSolver
  • Business Intelligence
    • Tableau
    • Qlik Sense
    • PowerBI
  • Big data
    • Spark (incl. Streaming)
    • HDFS
    • Hive
    • Kafka

Our Working Environment


As a group committed to diversity, data scientists and engineers from all backgrounds and different skill sets bring their expertise to our data science teams. Our international teams consist of physicists, statisticians, natural scientists, computer scientists and engineers from various fields. Each individual brings his or her own expertise to team, to come up with the best solutions.

Data Science Community

DPDHL Group employs many data scientists among divisions and global locations and we prioritize collaboration across all our divisions and location. We arrange community gatherings such as our regional Lunch and Learn meetings, where our scientists share project experiences, and our annual Data Science Day, where our data scientists get together in Bonn for a full day of internal and external presentations and information exchange. For more academic endeavors, some data scientists have frequent meetings in our Data Science Journal Club, where the topic is cutting-edge data science research.

External Collaborations

In addition to our internal community, we also have ties with external data science communities. We do occasional exchanges with the data science departments of other companies on methods and use cases. We also share our knowledge in public events like regional meet-ups.

Do you have an innovative idea for your data science thesis that relates to data analytics and the logistics industry? Then we look forward to hearing about it. We do occasionally have students working on their theses on-premise.

Continuous Learning

In this challenging world of competition, our data scientists are in full control of their own development – with the sparring and mentorship of our data science colleagues. We give all new joiners continuous onboarding sessions to learn about our past projects – from a business and technical perspective. To keep up with the latest advancements in both literature and technology, our teams participate in workshops and seminars all around the world. Furthermore, we conduct regular internal trainings and participate in external trainings where we see a need for improvement. It is absolutely vital for us that our data scientists and engineers continue to learn – for the sake of themselves and our company.

Our Application Process

We know that as a Data Scientist it might be hard to estimate your day-to-day tasks when you are applying for a position. That is why our interview process is designed to give you a glimpse of what we do, while understanding your capabilities and making sure we find the best matching position to you within the DPDHL Group.

  • Application
  • Telephone Interview
  • Interview Day


Application: You can apply through our job postings with your CV and a cover letter.
See open positions HERE

Telephone Interview

Telephone Interview: We introduce our company to you and let you introduce yourself. After a thorough talk about your CV, we walk you through one of our real-life use-cases and assess your business understanding and technical capabilities. Feel free to jump in with lots of questions!

Interview Day

Interview Day: A half-day journey in the Post Tower (or sometimes on Skype) to walk you through even more real-life projects. In your 2 or 3 interviews you meet more people from our teams and business partners, and have plenty of time to learn more about the group and/or target position(s).

To be completely honest, the most important things to use are not “where do you see yourself in 5 years?”, “what kind of music do you listen to?” or even “why should we hire you?”

We know that, 5 years from now, most data scientist want to work on even more challenging tasks – and some might even want to take over a leadership role. What is important to us is that we can help you develop in your desired direction.

We believe that a different taste in music brings diversity to our team – but that it’s unlikely to be an indicator of your competency with Data Science.

Finally, we assume that you wouldn’t be here in the first place; if you thought we shouldn’t hire you. :)

However, we might be asking you questions like:

  • What data-science projects would you want to pursue in DPDHL if you were given all resources you need?
  • What is your favourite machine learning model – and why?
  • How can you explain your thesis to a non-data scientist?
  • How do you make sure your model gets into day-to-day use within our warehouses?
  • How do you convince a transportation manager to use your model?

Data Scientists @ Deutsche Post DHL



Head of Department


Phone: (+49) 228 1896 3374