Applied Scientist - Transaction Risk Management in Berlin bei Zalando

At Zalando, our vision is to be the Starting Point for Fashion. We want to offer a shopping experience that is characterized by trust, for our more than 50 million customers in 25 markets across Europe and our +6,500 partner brands. To maintain this trust, it is vital for us to manage customer and purchase risks that originate from fraudulent behaviors on our fashion platform. With 3.3 million shopping items, resulting in hundreds of thousands of orders every single day, we use big data and advanced methods from machine learning to predict and mitigate such risks and ensure trustful relationships with our customers and partners.

As a full-stack applied scientist in our Transactions Risk Management team you will have the opportunity to join a dynamic and diverse group of engineers and scientists. As an analytics team, we are responsible for several predictive services to safeguard other teams in the checkout domain at Zalando. As part of our team, you will have the chance to work on cutting edge projects, raise the technical bar, improve our operational excellence, and shape our ways of working.

What you build and put in production is impacting not only every single Zalando customer on the spot, but also the performance of Zalando and its partners.

WHERE YOUR EXPERTISE IS NEEDED

• Take end-to-end ownership for developing, deploying, and operating machine learning solutions for detecting, predicting, and managing customer and purchase risks

• Quick prototyping and spiking of machine learning models to assess their applicability for solving research, customer, and business problems

• Tackle challenges for developing algorithms and running them efficiently on resource constrained platforms

• Monitoring and optimizing machine learning infrastructure running on AWS and Databricks/Spark

• Conducting (ad-hoc) exploratory analysis based on big (un-/semi-)structured data to discover new suspicious behaviors on our fashion platform

• Rigorous approach in solving, conducting, and documenting research projects

• Work closely with software engineers, applied scientists, data analysts, product managers, and fraud specialists to solve the problem of fraud

• Contribute to our growing science community and encourage knowledge sharing in an agile work environment

WHAT WE’RE LOOKING FOR

• 1-3 years of hands-on experience as an applied scientist, developing and productionizing machine / deep learning models in cloud environments (preferably AWS)

• Good proficiency in Python and related machine / deep learning frameworks, such as Pytorch, Tensorflow, Keras, etc.

• Expertise in machine learning infrastructure and tooling, such as Databricks, Spark, Flink, relational databases, AWS SageMaker, S3, EC2, Step Functions, Git

• Experience with data storage, ingestion, and transformation, also including machine learning workflow orchestration

• Passion for developing clean, well maintainable, and testable code

• Motivation for continued personal development in discovering new technologies and software services

• Ability and eagerness to understand the business context where the team operates and the customer problems being solved

• Good communication skills to translate (even complex) analytical / engineering decisions and outcomes to broader, non-technical audience

Preferred

• Previous knowledge in working with un-/weak-labeled data (self-supervised models, synthetic label generation)

• Experience in designing, developing, and operating highly-scalable microservices on a distributed system

• Knowledge in automated deployment and monitoring through CI/CD pipeline (Docker, Kubernetes, or similar)

• Work experience with a high level of test automation (unit, component, integration)

• Running and evaluating experimental machine learning deployments (canary, blue-green)

• Knowledge about machine learning on graphs, including community detection, graph embeddings, and graph neural networks.

PERKS AT WORK

• Culture of trust, empowerment and constructive feedback, open source commitment, meetups, game nights, 70+ internal technical and fun guilds, knowledge sharing through tech talks, internal tech academy and blogs, product demos, parties & events

• Competitive salary, employee share shop, 40% Zalando shopping discount, discounts from external partners, centrally located offices, public transport