Mika Rebensburg

Founding ML Engineer @ Tabula (since Jan 2026)

Science-informed, developer-driven ML engineering.

From an ICLR 2025 workshop paper to 100k+ app downloads, I focus on machine learning that turns strong research into reliable product outcomes.

See current focus Email me

Current focus

How I approach work

Production ML Systems

I build model-driven features that need to run reliably in real workflows, from prototyping to rollout and monitoring.

Founding ML Engineer @ Tabula · present

LLM featuresDeployment qualityReproducible pipelines

Scientific Machine Learning

I have worked on generative methods for molecular and graph settings, combining theory with practical implementation.

FreeFlow workshop paper · ICLR 2025

Flow matchingGenerative modelsMolecular systems

End-to-End Ownership

I like owning the full lifecycle: data, modeling, evaluation, and delivery.

Android app downloads · 100k+

Python/PyTorchExperiment designCross-functional delivery

Journey

From foundations to production ownership

  1. 10/2018-05/2022

    B.Sc. Computer Science

    Technical University Berlin

    Graduated with honors (top 3%).

    Education

  2. 08/2020-10/2021

    Study Abroad

    Lund University, Sweden

    Selected via ERASMUS+ exchange program.

    Education

  3. 07/2021-08/2023

    Working Student Machine Learning

    Exxeta

    Built and shipped a web-crawling ML pipeline to identify organizations aligned with UN sustainability goals.

    Experience

  4. 04/2023-11/2025

    M.Sc. Mathematics in Data Science

    Technical University Munich

    Passed with distinction (sehr gut).

    Education

  5. 09/2023-08/2024

    Student Research Assistant

    TUM Data Analytics and Machine Learning Group

    Researched generative models for graph and molecular modeling (PyTorch).

    Research

  6. 10/2023-11/2024

    MIT Project Collaboration

    MIT x TUM

    Flow matching for estimating free-energy differences between molecular systems; workshop publication (ICLR 2025).

    Research

  7. 09/2024-02/2025

    Applied Science Internship

    Amazon Development Center Germany

    Improved deep-learning and LLM solution performance for production systems.

    Experience

  8. 01/2026-Present

    Founding ML Engineer

    Tabula

    Building AI products for the tax domain with a focus on practical model impact.

    Experience

Featured build

Library App (Android)

Library App interface from the legacy portfolio

I started this app to learn Android development and turned it into a long-running product with sustained usage. It remains one of my clearest examples of end-to-end product ownership.

  • Designed, developed, and maintained the app independently since 2018
  • Reached 100k+ downloads through organic user adoption
  • Supports cataloging, wishlist management, search, and lending workflows
  • A long-term end-to-end product project from feature ideation to production maintenance
View on Google Play

Credentials

Core background and working toolkit

Education highlights

  • M.Sc. Mathematics in Data Science

    Technical University of Munich

    Passed with distinction (sehr gut)

  • B.Sc. Computer Science

    Technical University of Berlin

    Graduated with honors (top 3%)

Toolkit and profile

Programming
Python (PyTorch, NumPy, Pandas), Java, Kotlin
Languages
German (native), English (C1), Swedish (B1)
Interests
Generative models, graph ML, NLP, LLM agents, product-minded app development

Contact

rebensburgmika@gmail.com LinkedIn profile Impressum Datenschutz

Berlin, Germany