About
Audio Summary
Listen to the Summary of Johannes Profile by Google's NotebookLM.
Experience
Co-Founder & CTO
Raumdeuter GbR • Leipzig, Germany
Nov 2024 - Dec 2025
Building a modern communication and engagement platform for members and fans of clubs and associations to enable participation and engagement — a way to voice your opinion. I lead the technical development: from conceptualization and architecture to implementation of the product as an customer-deployed MVP with the help of a 4-person development team and coordinated external IT partners.
- Given an ambiguous initial problem configuration, I led the product ideation workshops to align team vision and product-solution-fit roadmap.
- Built engagement platform from 0 → 1 in less than 12 months on a €30k budget (likely resulting in approx. EUR 1.5M post-money valuation).
- Owned software architecture of the SaaS: Designed Microservices utilizing Docker Containerization (Distroless, Multistage Build) using robust and easily expandable Service Mesh patterns to ensure fast scaleability and prevent expensivley rebuilding 'prototype slug' later on.
- Stabilized the frontend delivery pipeline: configured and optimized Next.js builds in a multi-workspace monorepo, and ensured reproducible builds across environments to lower time spent+costs per dev on plattform issues.
- Established DevEx/DevOps: CI/CD, Linting and Pre-commit Hooks, logging and testing across Next.js and Python services, cutting release friction and significantly improving onboarding speed for new hires.
- Owned system administration and networking: Microsoft tenant administration, SSL/TLS/DNS management and nginx routing configuration for domains and services.
- Developed Machine Learning and (agentic) NLP pipelines in Python: Semantic Text Chunking + RAG vectorization using Qdrant, Sentiment Analysis, Topic Clustering and content summarization as analysis services and core value contributor to our customer.
- Established GDPR-compliant data privacy and Privacy by Design practices to address jurisdictional circumstances and obtain a valuable quality label.
- Owned end‑to‑end technical delivery and coordinated interns, contractors, and external partners; partnered closely with the frontend lead.
Mathematics Teacher (Substitute)
Rahn Education (Private School) • Leipzig, Germany
Nov 2022 - Jun 2023
Responsible for the supervision of talented pupils (Groups of 4 to 8)
- Prepared customized university level material to challenge and spark interest for Mathematics.
- Initiated a pupil-led lecture series about Artificial Intelligence and its use cases.
- Taught basic principles of the mathematics behind 3D printing and coding.
Student Assistant
Child and Adolescent Psychiatry, Philipps University Marburg • Marburg, Germany
May 2019 - Jan 2021
Responsible for the validation and reproduction of earlier results of research as well as for programming Machine Learning algorithms in R to solve classification problems in the field of ASD diagnosis.
- My analysis results supported the findings of 3 peer-reviewed publications (listed in publications) and have been cited ~70 times.
- Worked highly interdisciplinary with psychologists and medical doctors to understand the clinical background and implications of the data + explain the statistical methods to non-data-scientists.
Intern, Data Analyst
Westphalia DataLab GmbH • Münster, Germany
Aug 2019 - Oct 2019
As a Data Analyst intern at Westphalia DataLab GmbH, I worked on a medical-diagnostics project combining 3D imaging and clinical data; developed and improved an experiment-tracking R/Shiny frontend and performed clinical-data analysis with Python and R.
- Worked out and enhanced existing features for an experiment-tracking R/Shiny frontend to improve researcher workflow and usability
- Analyzed 3D-image and clinical datasets with Python and R to derive actionable diagnostic insights
- Integrated and managed project data using MongoDB and PostgreSQL and maintained code/versioning with Git
Intern, Sport and Athletes Marketing
Red Bull Germany GmbH • Munich, Germany
Mar 2016 - Aug 2016
In my role as a Sports and Athletes Marketing Manager, I updated and maintained the Athletes Portfolio as an asset of high value to the marketing mix of Red Bull. I furthermore participated in the Business Planning Process for 2017 by generating Ideas, their evaluation and processing for presentation
- Team Activation of THW Kiel including personalized Asset Design + Production, and a flight for the Team in one of Mateschitz's private hangar airplanes.
- Event Supervision and organizational Role in Projects such as the Hospitality Area for Red Bull's VIPS at MotoGP 2016.
Student Brand Manager
Red Bull Germany GmbH • Hamburg, Germany
May 2013 - Dec 2016
The Student Brand Manager Programm by Red Bull is designed find and educate the next generation of Red Bull Marketing Experts. As a Student Brand Manager, I was responsible for organizing guerrilla marketing events and creating brand awareness among the student community. The role as a SBM allowed me to develop strong marketing and leadership skills while gaining practical experience in brand management from one of the most recognized brands worldwide.
- Student Brand Manager of the Year 2016 in North Germany
- Organized own campus events and moderated campus events of other Student Brand Managers
- Participated and supported regional and national events such as Red Bull Seifenkistenrennen and Red Bull Coast2Coast
Working Student, Human Resources
Goodgame Studios • Hamburg, Germany
Aug 2011 - Sep 2012
Starting out as an intern and later becoming a working student, I supported the HR team in various administrative tasks, recruitment processes, and employee onboarding activities. This role provided me with valuable insights into HR operations within a fast growing startup company.
- The HR recruiting process was responsible for hiring 70 new employees (from 30) within one year, contributing to the company's rapid growth.
- I was responsible for scouting and selecting candidates, scheduling interviews, and coordinating communication between applicants and hiring managers, improving the efficiency of the recruitment process.
Projects
AutoCoverletter
Agent-assisted cover letter automation
Automates the end-to-end cover letter workflow using agents, local MCP resources, and context engineering to keep prompts focused while staying aligned with local application data.
- Automated about 90% of the cover letter writing process, including Notion sync and application tracking
- Strong coverletters without flattering, lying or exaggeration
- Neat design and consistency
- Achieved 10%+ invitation-to-call rate during a weak hiring market
Raumdeuter
Fanrelationship Management Software
Listen. Understand. Act. The Raumdeuter Software helps Sportclubs turn fan opinions into data-driven decisions and sustainable knowledge
- Raumdeuter stands for the intuition to recognize and transform opportunities, like Thomas Müller on the field. The software provides sports clubs with the tools to translate their fans' voices into successes off the field.
- Quantitative surveys, summarization of similar texts using embedded vectors and clustering methods in qualitative surveys, sentiment analysis of results for early trend detection
- The software integrates seamlessly into existing systems such as websites, club apps, or member portals. This keeps everything within the familiar environment of the club. Fans do not need to learn a new platform but can participate directly where they are already active.
- Pilot project (website integration + users) with a football club from the region (Chemie Leipzig)
Neural Machine Translation with Pivot Strategies
Master's Thesis on Low-Resource Language Pairs
Master's thesis demonstrating that pivot translation strategies can achieve compareable translation quality with 99% less parallel training data, making NMT viable for low-resource language pairs.
- Achieved 70-80% of direct translation quality using only 1% of parallel data by leveraging English as a bridge language between French and German
- Best-performing pivot model showed substantial improvements over baseline: +15.4% BLEU, +5.8% chrF, and -4.7% TER scores
- Demonstrated clear relationship between pivot resource availability and translation quality using transformer architecture
- Provided fundamental insights for low-resource NMT and large language model development
Cook Up Kitchen
PopUp Restaurant
A few friends and me started a repeating PopUp Restaurant in Hamburg to explore whether food blogging could be a viable marketing strategy for restaurants, which resulted in a since then lasting business opportunity for one of us.
- 2nd Event: Video produced by professional videographer available on Vimeo
- 3rd Event: catered to 100 people in one night on a 3-course menu basis in a kitchen very much not designed for that purpose
Publications
Phenotypic differences between female and male individuals with suspicion of ASD
Sanna Stroth, Johannes Tauscher, Nicole Beyer, Charlotte Küpper, Luise Poustka, Stefan Roepke, Veit Roessner, Dominik Heider, Inge Kamp-Becker • Molecular Autism, 13 (1), 2022 • 16 citations
Abstract
Although autism spectrum disorder (ASD) is a common developmental disorder, our knowledge about a behavioral and neurobiological female Phenotype is still scarce. As the conceptualization and understanding of ASD are mainly based on the investigation of male individuals, females with ASD may not be adequately identified by routine clinical diagnostics. The present Machine Learning approach aimed to identify diagnostic information from the Autism Diagnostic Observation Schedule (ADOS) that discriminates best between ASD and non-ASD in females and males. Random Forests were used to discover patterns of symptoms in diagnostic data from the ADOS (modules 3 and 4) in 1057 participants with ASD (18.1% female) and 1230 participants with non-ASD (17.9% % female). Predictive performances of reduced feature models were explored and compared between females and males without Intellectual Disability. Reduced feature models relied on considerably fewer features from the ADOS in females compared to males, while still yielding similar classification performance (e.g., Sensitivity and Specificity). As in previous studies, the current sample of females with ASD is smaller than the male sample and thus, females may still be underrepresented, limiting the statistical power to detect small to moderate effects. Our results do not suggest the need for new or altered diagnostic algorithms for females with ASD. Although we identified some Phenotype differences between females and males, the existing diagnostic tools seem to sufficiently capture the core autistic features in both groups.
Is the Combination of ADOS and ADI-R Necessary to Classify ASD? Rethinking the “Gold Standard” in Diagnosing ASD
Sanna Stroth, Johannes Tauscher, Nicole Beyer, Charlotte Küpper, Luise Poustka, Stefan Roepke, Veit Roessner, Dominik Heider, Inge Kamp-Becker • Frontiers of Psychiatry, 12, 2021 • 45 citations
Abstract
Diagnosing autism spectrum disorder (ASD) requires extensive clinical expertise and training as well as a focus on Differential Diagnosis. The diagnostic process is particularly complex given symptom overlap with other mental disorders and high rates of co-occurring physical and mental health concerns. The aim of this study was to conduct a data-driven selection of the most relevant diagnostic information collected from a behavior observation and an anamnestic interview in two clinical samples of children/younger adolescents and adolescents/adults with suspected ASD. Via Random Forests, the present study discovered patterns of symptoms in the diagnostic data of 2310 participants (46% ASD, 54% non-ASD, age range 4–72 years) using data from the combined Autism Diagnostic Observation Schedule (ADOS) and Autism Diagnostic Interview-Revised (ADI-R) and ADOS data alone. Classifiers built on reduced subsets of diagnostic features yield satisfactory Sensitivity and Specificity values. For adolescents/adults specificity values were lower compared to those for children/younger adolescents. The models including ADOS and ADI-R data were mainly built on ADOS items and in the adolescent/adult sample the classifier including only ADOS items performed even better than the classifier including information from both instruments. Results suggest that reduced subsets of ADOS and ADI-R items may suffice to effectively differentiate ASD from other mental disorders. The imbalance of ADOS and ADI-R items included in the models leads to the assumption that, particularly in adolescents and adults, the ADI-R may play a lesser role than current behavior observations.
Identification of the most indicative and discriminative features from diagnostic instruments for children with autism
Sanna Stroth, Johannes Tauscher, Nicole Wolff, Charlotte Küpper, Luise Poustka, Stefan Roepke, Veit Roessner, Dominik Heider, Inge Kamp-Becker • JCPP Advances, Volume 1, Issue 2, 2021 • 9 citations
Abstract
Diagnosing autism spectrum disorder (ASD) is complex and time-consuming. The present work systematically examines the importance of items from the Autism Diagnostic Interview-Revised (ADI-R) and Autism Diagnostic Observation Schedule (ADOS) in discerning children with and without ASD. Knowledge of the most discriminative features and their underlying concepts may prove valuable for the future training tools that assist clinicians to substantiate or extenuate a suspicion of ASD in nonverbal and minimally verbal children. In two samples of nonverbal (N = 466) and minimally verbal (N = 566) children with ASD (N = 509) and other mental disorders or developmental delays (N = 523), we applied Random Forests to (i) the combination of ADI-R and ADOS data versus (ii) ADOS data alone. We compared the predictive performance of reduced feature models against outcomes provided by models containing all features. For nonverbal children, the Random Forests classifier indicated social orientation to be most powerful in differentiating ASD from non-ASD cases. In minimally verbal children, we find language/speech peculiarities in combination with facial/nonverbal expressions and reciprocity to be most distinctive. Based on Machine Learning strategies, we carve out those symptoms of ASD that prove to be central for the differentiation of ASD cases from those with other developmental or mental disorders (high Sensitivity and Specificity in minimally verbal children). These core concepts ought to be considered in the future training tools for clinicians.
Education
M.Sc in Data Science, Faculty of Mathematics and Computer Science
Leipzig University • Leipzig, Germany
Apr 2021 - Oct 2024
- Overall Score: 1.7, personal focus on advanced Statistics, Machine Learning, Data Privacy, LLMs
- Master's Thesis (1.1): Neural Machine Translation with Transformers - Leveraging the Pivot Technique for Low-Resource Language Pairs
B.Sc. in Data Science, Faculty of Mathematics and Computer Science
Philipps University Marburg • Marburg, Germany
Oct 2017 - Mar 2021
- Bachelor's Thesis (1.7): Multidimensional Data Exploration and Visualization of Membrane Proteins Attributes
- Curriculum focusses on Mathematics, Statistics, Computer Science and Programming
B.A. in Marketing and Technical Business Administration
HAW Hamburg • Hamburg, Germany
Mar 2012 - Oct 2017
- Bachelor's Thesis (1.7): Use of Linear Discriminant Analysis as an alternative to tree analysis - methodological evaluation and exemplary implementation using a survey on the success of movies.
- Curriculum focusses on Marketing, Engineering and Business Administration
Skills
- Raumdeuter GbR
- Westphalia DataLab GmbH
- Child and Adolescent Psychiatry
- Master's Thesis
- Bachelor's Thesis
- Child and Adolescent Psychiatry
- Westphalia DataLab GmbH
- Raumdeuter GbR
- Master's Thesis
- Linux System Administration
- Linux
- Introduction to Computer Science and Software Engineering during Bachelor's Degree
- Raumdeuter GbR
- Westphalia DataLab GmbH
- Bachelor's Thesis
- Raumdeuter GbR
- Github Page
- The Odin Project
- Raumdeuter GbR
- Master's Thesis
- Raumdeuter GbR
- Westphalia DataLab GmbH
- Raumdeuter GbR
- Westphalia DataLab GmbH
- Raumdeuter GbR
- Raumdeuter GbR
- Raumdeuter GbR
- Red Bull Germany GmbH
- Hobby Projects
- Raumdeuter GbR
- everywhere
- Raumdeuter GbR
- Github Page
- Child and Adolescent Psychiatry
- Master's Thesis
- Bachelor's Thesis
- Bachelor's Thesis (Economics)
- Westphalia DataLab GmbH
- Child and Adolescent Psychiatry
- Tutoring
- Economics Bachelor's Thesis
- Westphalia DataLab GmbH
- Raumdeuter GbR
- Master's Thesis
- Raumdeuter GbR
- Master's Thesis
- Interests & Hobby
- Child and Adolescent Psychiatry, Philipps University Marburg
- Child and Adolescent Psychiatry, Philipps University Marburg
- Westphalia DataLab GmbH
- Master's Thesis
- Raumdeuter GbR
- Raumdeuter GbR
- Red Bull Germany GmbH
- Goodgame Studios
- Raumdeuter GbR
- everywhere
- Raumdeuter GbR
- Westphalia DataLab GmbH
- Master's Thesis
- Bachelor's Thesis
- Child and Adolescent Psychiatry, Philipps University Marburg
- everywhere
- Raumdeuter GbR
- Rahn Education (Private School)
- Master's Thesis
- Bachelor's Thesis
- Westphalia DataLab GmbH
- Child and Adolescent Psychiatry, Philipps University Marburg
- Raumdeuter GbR
- Westphalia DataLab GmbH
- Master's Thesis
- Bachelor's Thesis
- Child and Adolescent Psychiatry, Philipps University Marburg
- Master's Thesis
- Bachelor's Thesis
- Academic Papers
- Raumdeuter GbR
- Raumdeuter GbR
Hover over a category for details
Programs & Awards
Programs
exist Startup Grant
2024-2025
Developed and launched a startup idea as part of the exist scholarship program funded by the German Federal Ministry for Economic Affairs and Climate Action (BMWK).
- Winner of the Impulse Summit 2025.
- Winner of the TGFS Technologiegründerfonds Sachsen Award as part of the HHL Digitale Space Program 2025.
- Winner of TechStart Dresden 2025.
- Top 5 out of 150 finalists at Samsung: Solve for Tomorrow 2025.
- Visited selected business seminars on diverse topics from Leipzig University and Leipzig Graduate School of Management (HHL).
Awards
DataFest 24h Hackathon
2019
Developed a collision detection algorithm and visualization for data from a season of women's rugby.
- Won 3rd place out of 19 teams for both best data insight and best data visualization.
Get in Touch
I'm currently looking for new opportunities. Feel free to reach out!