Nicolas Schönborn

For a local small-sized metalworking company, I've developed multiple tools to automate recurring manual work. This includes automated order communication between the ERP and production systems as well as automated feedback about the current production status. Additionally, a React Native application was developed, which provided a time tracker for the employees allowing them to track detailed information about specific orders. It included quality assurance functionalities allowing employees to capture production results. The app also contains a storage tracking tool that enables employees to add sub-assemblies and products to storage units by using an in-app barcode scanner. The back end was implemented in Node.js. A web presence has also been designed and developed.

I developed a prototyping tool for quickly creating web applications with a component-based design tool. The goal is to reduce time to market and make web development accessible for non-developers. The application is implemented with React and creates React applications.

A Magento based eCommerce shop for homeware products that features multiple unique instances for end customers, business customers, and internal employees. The shop was taken over from an agency that failed to deliver a final product. I cleaned up the project, added new features, brought the shop to production, and gave it to another Magento-focused agency, which continued managing and enhancing the shop.

A Shopware based B2B eCommerce shop for a sanitary and heating construction company that digitalized their ordering process completely for their business customers. The shop included individual plugins providing the ability for Arvendo to address specific customers with individual offers. A requirement was to allow a single company to have different users who could create multiple order lists; the companies' purchaser could then decide about the purchases individually by submitting an order list. The customers could print QR codes to attach them to their storage units and directly find the product in the shop to order via their mobile devices.

In a student project together with a major German car manufacturer, my team and I developed a system around the safety of drivers of autonomous taxis and their fleet managers. I led the team and was responsible for the design and development of the app written in React Native and MobX. The app included all the tools required during the workday of a safety driver. It started by communicating to the driver his working hours, guiding him to the autonomous vehicle for his current shift, doing prior safety checks (including a visual tracking system) before starting the ride, the management of the vehicle, and the notification of possible incidents during the ride.
The fleet manager's web app was implemented in Angular and allowed to manage the taxi fleet, safety drivers, their working hours, and their assignment to cars. The fleet manager could always see where the autonomous cars are, inspect current events, including major incidents.
The back end was written in Spring and was initialized with JHipster. We used Elasticsearch to implement logging of events and, in case of incidents, the communication of those events.

Within a team of five, we trained a DNN to play the game of Rummikub with reinforcement learning. It was challenging due to the game's vast action space, which is a hot topic in current research and the implementation of a multi-agent concurrent learning system. The game itself was developed in React. The game engine was implemented in Python, and the DNN was trained with PyTorch. My responsibility was the design and planning of the game engine in a multi-agent learning system. I also oversaw the implementation of the game's graphical interface, the provision of the learning environment on Ubuntu, and the execution of additional experiments.

A prototype app that allows users to synchronously play music, videos, and sounds on multiple devices by taking advantage of the W3C Multi-Device Timing API. The app was implemented in React Native. For the back end, I used Node.js and MongoDB. The app was developed as a student project at TU Berlin together with Fraunhofer FOKUS.

An early-stage technical prototype of a Bitcoin exchange service in the early days of Bitcoin. I developed the real-time trading engine, user, and wallet management, which included withdrawal and deposit of Bitcoins, the user interface featuring a simple trading view with live price updates, and a candlestick chart. I was guided by other developers about application security and event-driven databases, and learned a lot about web security in a short period of time.

Developed a flappy bird clone implemented in Unity 3D and published into the Google App Store. It reached over 500 downloads but is currently not available anymore. The graphics were created with Inkscape.