Manuel Oelmaier

About me

Computer Science student at the Technical University of Munich. Experienced in building full-stack applications, primarily leveraging Python, Java, Webstack and adept at quickly adopting new technologies. Specializing in Machine Learning and AI , with hands-on experience in fine-tuning transformer models and model evaluation. Passionate about developing solutions that bridge technical innovation with real-world impact. In my free time I enjoy playing Chess, reading books and being active in the Gym.

Principles

Pragmatism

I approach software development with a balance of aiming high and being realistic . My priority is delivering real, tangible value within realistic timeframes. My focus is on pragmatic iteration: choosing simple solutions that just work, only adding complexity when it genuinely pays off. This means I prioritize new features and improving with user feedback, over chasing perfection.

Software Craftsmanship

With a strong foundation in computer science, I’m passionate about building reliable, scalable, and thoughtful software. I enjoy working across the stack, whether it’s crafting clean backend architectures, designing intuitive interfaces, or integrating the latest ML/AI models. As I grow as a developer, I strive to write maintainable, human-friendly code that empowers others and leads to meaningful, high-quality products.

Lifelong Learning

Technology changes fast, and staying curious is my engine for growth . I continually explore new tools, frameworks, and ideas—whether through building side projects (like ManuChess) or studying academic concepts (like LLMs and transformer architectures). For me, learning isn't a goal but a hands-on habit , ensuring I'm constantly adapting and turning new knowledge into practical, high-impact skills.

Résumé

Download my résumé for a concise overview of my skills, experience, and projects

Grab A Copy

Project Showcase

Prototype: ManuChess

A small website which lets you play against a Mirror of Myself. I am currently working on training a small Neural Net with all my Chess games (tournament and online), which at the end will play like me. Coming soon!

View on GitHub

Webstack

MongoDB

Pytorch

LocalPythonCodingLlm

As Coursework, we fine-tuned and compared different small LLMs in their Ability to write Python Code. We made the best one publicly available via a VSC Extension, where the model can be queried locally. No more code freely shared to LLMs.

View on GitHub

Python

Pytorch

Typescript

Quiz App: Theoretical Informatics

A small website with the Collection of the Past Multiple Choice questions of public Exams, to help students prepare. Renders Formulas and Mathematical Expressions accurately.