Kevin Schestakow

I was able to learn JavaScript at university during my bachelor's degree (Software Development I+II) and further improve it at the Fraunhofer Institute for Building Physics (IBP) as a working student. My bad habit is that I use it for almost everything.

I first encountered TypeScript during my bachelor's degree while working on larger projects. Suddenly, I had strings where I expected numbers and vice versa. At Fraunhofer IBP, I was able to deepen this knowledge significantly when I worked on a large, legacy codebase.

It's fascinating what Node.js enables developers to accomplish. An indispensable tool that I work with almost daily as an effective solution.

Next.js has become my go-to for building scalable, high-performance web applications. The component model, server actions, and the massive ecosystem make it an absolute joy to work with.

Vue.js and Nuxt.js hold a special place in my heart. I have extensive experience building complex applications with them and love the elegance and simplicity they offer for rapid development.

TailwindCSS is my styling engine of choice. Whether I'm working in React or Vue, Tailwind allows me to build beautiful, responsive designs with incredible speed and consistency.

I prefer Express.js for its intuitive syntax and efficiency in rapid development of web applications and APIs. It's particularly suitable for building lightweight and scalable backend solutions.

As part of my Master's program in Digital Processes and Technologies, I'm currently expanding my expertise in Java and Spring Boot backend development. While I'm still building and consolidating my skills in this area, I find it more challenging compared to Python, JavaScript, or Node.js backends.

Flask is my go-to Python web framework for building lightweight web applications and RESTful APIs. I appreciate its minimalist approach and flexibility, making it perfect for rapid prototyping and small to medium-sized projects.

I'm passionate about automation and deeply interested in LLM models, machine learning, and AI development. Python is indispensable in these fields, despite JavaScript alternatives like TensorFlow.js. I frequently work with Hugging Face models, where Python excels in terms of simplicity and efficiency. Python was my second programming language, learned before Java and Spring Boot. For my bachelor's thesis at Stuttgart Media University, I focused on GPT fine-tuning and developed a GPT-powered coding assistant plugin for students. This project involved model training, AI analysis, statistics, and Weights & Biases implementation, all executed in Python.

Even simple scripts can greatly streamline workflows. During my 6-month internship at Mercedes Benz, I developed several large-scale worksheets and VBA scripts to automate solutions, improve efficiency, and save time. While this involves more scripting than traditional programming and is not my primary focus, I found it fascinating how simple scripts could save minutes of daily work. VBA is a powerful tool, especially in MS Office applications like Excel and PowerPoint. In my experience, Excel is where VBA truly shines, enabling implementation of sophisticated automation solutions.

Bootstrap is excellent for its simplicity and particularly beginner-friendly. While I started my journey with Bootstrap, I have now primarily transitioned to Tailwind CSS, unless a project specifically requires Bootstrap implementation.

Supabase is my preferred open-source alternative to Firebase. It provides a complete backend solution with PostgreSQL database, authentication, real-time subscriptions, and storage - all with a developer-friendly API that makes building modern applications fast and straightforward.