I am a data-driven problem solver with years of experience in AI, Data Engineering and Machine Learning. I have strong technical skills in Mathematics, Data Science, and Software Engineering. I'm driven by the desire to work in growth environment and committed to make a meaningful impact on the world through AI. I'm a team player with a strong "get-it-done" attitude and a passion for continuous learning.

Educational Foundation & Skill Development

  • Master’s degree in Data Science from the Technical University of Munich, preceded by a Bachelor’s in Mathematics.
  • Six years of diverse practical experience in Product Building at Pyne, Machine Learning at Genesis and Mergespot, Software Engineering at Microsoft and Tacto, and Research at Telecom Paris.
  • Deep interest in NLP, Speech Recognition, and Knowledge Graphs, with notable research under Prof. Thomas Bonald at Telecom Paris.

Professional Impact & Innovations

My latest and biggest impact on users comes from my work at pyne.ai. I have joined the team as a Founding Engineer. I contributed to the team with knowledge on how to build MVPs quickly, iterate ideas fast and keep the product at the centre of everything.

Product #1 - Building a Revenue operating system for B2B SaaS companies

  • Developed data analytics platform for RevOps teams.
  • Designed data model, built ETL pipeline and integrated with Hubspot, Salesforce and Chargebee.
  • Created Insights system using ML and LLMs: GPT-4 with Vision + diagrams + tabular data = insights.
  • Responsible for diverse functions: Data Management, Backend Development, Product & Hiring.

Product #2 - Guided product tours

  • Co-Developed a browser-based extension that captures user interactions and voice, processes data using Speech Recognition and LLMs, and creates interactive product tours available as video, chatbot, or web page injection.
  • Developed a ChatBot for marketing teams based on RAGs.
  • Designed quality evaluation and retraining pipeline for RAGs.

It is also worth mentioning my work at Microsoft and contributions to the Azure Speech Service product. Due to my curiosity, I have worked in multiple teams.

At the Research team

  • I was responsible for training an Acoustic Diarist Identification system for the Swiss language.
  • I learned how to use the Kalidi framework, as well as how to run distributed training on multiple GPUs.

At the Speech Recognition team

  • I developed a pipeline for automatic deployment, testing and updating of Speech Recognition models available in more than 70 locales worldwide and serving 100k people daily.
  • This pipeline reduced the deployment time of new models into production by a factor of 10. I integrated the project into a vast codebase requiring communication with more than 10 Product Owners.

Looking ahead

My professional path is driven by a passion for continuous learning and a "get-it-done" attitude. As I continue to explore and contribute to the ever-evolving field of AI, my goal remains to harness technology to create meaningful and impactful solutions.