About

Short bio, research focus, and a bit of personal context.

Short Bio

I'm a Computer Engineering student at RWTH Aachen University and a Werkstudent Researcher at the Enterprise Information Systems (EIS) department of Fraunhofer IAIS. My work sits at the intersection of machine learning systems, agentic workflows, and embedded/edge AI.

  • RWTH CE student, focusing on ML systems and signal processing.
  • Werkstudent Researcher @ Fraunhofer IAIS (EIS).
  • Interested in making ML agents robust enough for real-world use.
RWTH Aachen Fraunhofer IAIS ML Systems

Research & Interests

I like approaching ML problems with a signal-theoretic mindset: treating logits, activations and time series as signals that can be analysed, regularized and clustered.

My first preprint, “Wavelet-Based Feature Extraction and Unsupervised Clustering for Parity Detection: A Feature Engineering Perspective”, explores how wavelet features and clustering can recover symbolic structure (odd vs. even) from purely numerical signals.

Current interests include:

  • Signal-theoretic metrics for LLM logit stability.
  • Wavelet-based feature extraction for structured signals.
  • Agentic systems and evaluation, especially agent-in-the-loop setups.
  • Edge and embedded ML deployments.

Talks

DataTalks.Club webinar (Dec 16, 2025)
“From Human-in-the-Loop to Agent-in-the-Loop: How ML Workflows Evolve into Agentic Feedback Systems”

A practical overview of how human feedback loops in ML pipelines turn into agent-in-the-loop systems: what changes in planning, tool use, evaluation, and how to keep systems debuggable as automation increases.

Personal

Outside of research and coursework, I enjoy working with electronics, microcontrollers and embedded systems—often building small tools that support my ML experiments. I also like documenting projects in public through blog posts and talks.

I'm especially motivated by projects where algorithms meet hardware and where “nice-in-theory” ideas have to survive noisy, real-world inputs.