Thomas Soenen

Designed and developed a user-customizable, distributed management, and orchestration (MANO) framework as the core element of the SONATA NFV (Network Function Virtualization) platform. The framework manages and controls the end-to-end lifecycle of virtual network functionality embedded in VMs and containers on top of cloud resources managed through Openstack and Kubernetes.
The MANO Framework consists of eight microservices, each embedded in a Docker container, which communicate using a RabbitMQ message bus; all microservices functionalities were written in Python. The application handles requests from customers to deploy and maintain their virtual network functionality. The MANO framework calculates the optimal host for this network functionality out of the available cloud resources and interacts with the appropriate cloud manager to establish deployment. The MANO framework interacts with software-defined networking managers to steer traffic from customers through their VMs or containers.
After the virtual network functionality is operational, the MANO framework monitors its quality of service and scales, migrates, reconfigures, or terminates the VMs and containers if required.

A Python library and CLI tool that consumes the REST API of an NFV Platform (SONATA NFV) and provides an easy-to-use CLI interface. The CLI tool builds on top of a python library to be used within Python as well. The tool significantly simplifies interfacing with the platform.

Designed and implemented an ETL pipeline To support the data scientists at TickUp AB that scrapes a third-party REST API every 10 minutes, keeping a PostgreSQL database up to date with the latest ticker information, which serves as input data for ML evaluations.

I had ownership throughout the entire process: designing the ETL, implementing it as a Python process, packaging it as a Docker container, creating a CI/CD workflow using Jenkins to test and deploy the container, and executing the process every 10 minutes through a Jenkins job.

Created a Python package for a BASF research group to facilitate the interaction of our data scientists with our software solutions built to support them. The Python package provides wrappers for the REST APIs of our various applications, making interacting with them easier for data scientists with a limited programming understanding. Among other features, using this package abstracted the authentication process for users, provided payload validation, and made for easy environment selection.