As data volumes continue to grow for eCommerce companies and the number of data consumers within the organization is increasing, sometimes old infrastructure will not be able to keep up with the challenges. Additionally, In this particular case, the computing and warehousing cluster has to be on-premise for data security reasons. After new cluster-infrastructure had been provided by an external provider, all data warehouse and computing logic has to be migrated from the old infrastructure to the new infrastructure. Additional challenges are maintaining backwards compatibility of the migrated processes at all times and adhering to strict security standards.

• Migration and Deployment of 30+ airflow DAGs with 20 ? 50 Tasks each on new infrastructure
• Co-development of a python client library for Apache Livy that is used by 100+ airflow tasks
• Deployment of 20+ Apache Hive databases with 10 ? 50 tables each in three Data Warehouse layers via Ansible
• Code review of 5-10 merge requests per week

In order to prevent financial and reputational loss in eCommerce platforms an automated security system is needed that can detect fraud patterns in online shop. The software, written in Java with Apache Flink, should be able to scale out over multiple shop systems and data sources. Further requirements are monitoring traffic in real time and incorporating expert knowledge alongside machine learning and Artificial Intelligence (A.I.) models. The software is deployed and operated on the customers cloud environtment by using modern Continuous Integration (CI) and DevOps principles.

• Lead design of the platform, Technical Lead for a team of 5 Developers
• Implementation of a proof of concept with Java from which 80% of code made the first product iteration
• Prototyping of two end-to-end MLOps workflows with MLflow and AWS Sagemaker
• Successful deployment and zero downtime operations on customer premises at around 15 million events per day
• Design of cloud based testing environment that can be brought up in less than 15 minutes (Infrastructure as Code) and handle up to 10 times of the production workload

A datadriven company needs to have a reliable and scalable infrastructure as a key components of the corporate decision making. Engineers as well as analysts need to be enabled to create ETL-processes, Artificial Intelligence (A.I.) jobs and ad-hoc reports without the need to consult with a data engineer. The data architecture of the company needs to provide scalability, clear separation between testing and production and ease of use. Modern DevOps practices like Continuous Integration (CI) and Infrastructure as Code need to be employed across the whole infrastructure.

• Leading conception of cloud based infrastructure based on the above requirements
• Initial training of 5 developers an onboarding of more than 10 developers since
• Initial setup and operation of apache airflow with intially ca. 10 jobs, scaling up to more than 100 regular scheduled jobs at present

In order to enable an eCommerce organization to become a datadriven organization there must be (among other things) a framework present to compare different version of the website against each other. Many members of the organization and departments need to be able to create and conduct experiments without the assistance of a data engineer. Anther important factor for the framework was the usage Bayesian statistics.

• Leading Conception of testing framework including randomization logic, statistical modelling and grapical presentation in the frontend
• Implementation of proof of concept for statistical engine
• Implementation of production code for frontend, backend, statistical engine
• Training of stakeholders from 3 different departments in methodology and statistical background of A/B-Testing

For an eCommerce Platform it is crucial to have a detailed picture of customer behaviour on which business decisions can be based. Either in real-time or from the data warehouse. For that a flexible, scalable, and fieldtestet solution is necessary which can run in the cloud. Additionally, all browser events need a custom enrichment with business information from the backend in order to provide necessary context e.g. for ?Add to Cart?-events. The webtracking pipeline is managed by using modern DevOps principles: Continuous Integration (CI), zero downtime deployments and Infrastructure as Code.

• Integration of snowplow event-pipeline in cloud based shop architecture
• Day to day operations of event-pipeline at ca. 4 million events per day
• Co-Engineering of custom enrichment in the webshop backend (ca. 1000+ lines of code) and handover of ownership to the backend team
• Setup of custom real time event monitoring (< 1s latency) with elasticsearch and kibana
• Setup of custom scheduling and deployment processes for 5 components of the snowplow event-pipeline

To enrich the shopping experience of the customer and to drive additional sales, the eCommerce platform should be able to recommend customers additional products. Two orthogonal strategies are employed: Product based similiarity based on neural network embeddings and collaborative filtering based on user behaviour. Additionally, Performance monitoring for the recommendations is needed.

• Productionize both models based on proof of concepts by ML engineer including data aquisition, running of the model and data output
• Scheduling and operations of productionized models, including 3 different code bases and more than 5 regularly scheduled jobs
• Operationalization of 10+ performance metrics over 5 dashboards for stakeholders