Michael Janz

-Technical lead for up to 15 people -Combination of technical lead and technical product owner -Defining of vision and development strategy -Designing deep learning solutions -Designing production ready machine learning models -Fields: NLG, NLP, NLU, Generation, Classification, Fact checking including pretraining, fine-tuning, using self-trained, open source and closed models as chatGPT or GPT-4

1. Pretraining and Finetuning of multilingual seq2seq Language models with the task of rewriting text in the domain of news. 2. Creation of german fact-checking Model. 3. Building infrastructure for handling large scale model-training 4. Recruiting, interviewing and onboarding 5. Deputy of Head of Machine learning. 6. Projekt management. Creating the roadmap for next ML models 7. Being part of setting up the organisation structure. 8. Technical Lead for fields including up to 6 Persons

At frogs42 I am responsible for developing and automatically evaluating new AI-models for conditional fact-based text generation. This includes all steps from researching new techniques in scientific papers to preparing new training data, training/finetuning a deep-learning model, automatically determine the quality of the model output and to put all these steps into software projects, that make the whole usage easy to handle.

At the KI Group, my Job is to analyze data and create workflows of very different kind. This includes working with Azure, like the azure cognitive service, analyzing data with python, and writing software with C#. I worked on a project for a big automobile manufacturer with the goal of building and comparing mathematical models for creating artificial customer trips. This included the use of markov chains, but also using neural networks (VAEs) for generating the trips.

Full stack developer on part time. I was in a team, responsible for developing the main product of Bayard Consulting. My main tasks were: - Fixing bugs and implement new features into our frontend and backend (AngularJs, Ruby, JavaScript, Html, Css) - Redevelop the testing environment (Cucumber, Cabybara, Ruby, Selenium) With these technologies, I also learned more about using Git, Maven, and Docker

I was in the agile development team of the Generali Deutschland Informatik Services GmbH. The mission of this team was to teach the company the basics of agile workstyle. My tasks were: -translating articles from english to german and vice versa -developing and designing a wiki for these articles -creating plans for the practical execution of the workshops

Because of a technical upgrade, existing data in Salesforce had to be translated into a new format. My task was, to do this translation. After planning the procedure, I wrote a script for translating the data in Salesforce. After a manual check, the task was successfully finished

At Klarna I gained more Experience in working with data, basically in Excel. My Job was to do calculations, create reports and analyze Data as part of the sales steering team.

Computer Reparaturen, Beratung zu Problemen, Professionelle Beratung bei Problemstellungen

Projekt-Management Assistenz in der IT • Entlastung des Projektmanagers durch Abnahme von Arbeitspaketen • Koordination von Reisen, Meetings und Termin • Ansprechpartner für Außenstellen • Allgemeine Bürotätigkeiten • Entwurf von Dokumenten und Plänen

Dies war der Zeitraum, indem unser Abschlussprojekt statt fand. Mein Team bestand aus 3 Leuten. Unsere Aufgabe war es, eine innovative Eingabesteuerung mit Windows Kinect zu programmieren, die Kamera soll mit Hilfe einer beweglichen Schiene immer den Benutzer im Fokus halten. Das Projekt war ein voller Erfolg, es konnten Eingaben per Bewegung getätigt werden, welche vorher definiert werden müssen und die Kamera folgt dem Benutzer.

IT-Support für externe Kunden und interne Kunden.

Physik Mathematik

Informationskompetenz Datenbank sowie Programmierung Wirtschaft Sprachwissenschaft Statistik Empirische Forschung

Programmierung: C# Netzwerk- und Rechnertechnik Best Practise