Ludwig von Schoenfeldt

Experienced in full-stack software development. Currently leveraging deep learning research in bioacoustic classification using CNNs, Transformers and RNNs. Interested in optimizing low-level architectures for ML/DL classification by exploring model distillation, pruning and quantization. Bilingual in English and German.

BS Computer Engineering, University of California San Diego (Expected June 2026)
Activities and societies: Focusing on ML research and hardware optimizations, ACM, IEEE

• Software Engineering Intern, Apple Inc., San Diego, US (June 2025 — Sept 2025)
  • Incoming Software Engineering Intern for this summer
• Undergraduate Student Researcher, Engineers for Exploration & KRG @ UC San Diego (Oct 2023 — Present)
  • Project Lead managing and training a team of 15 undergraduate students for various tasks (including ML, SE, DS)
  • Using CNNs to classify bird sounds in the Amazon rainforest in Peru in collaboration with the San Diego Zoo
  • Conducting advanced data science research with several terabytes of audio data on HPC using Pytorch, TensorFlow, ONNX, etc.
  • Exploring new approaches using Transformers, RNNs and hybrid models such as CNN Transformer Hybrid HuBERT
  • Developing Pytorch + Rust implementation to drastically speed up inference and training times for larger models
  • Developing a Desktop app using Electron.js which encapsulates labeling, running inference and storing data using SQLite
  • Awarded SRIP under mentorship of Professor Dr. Ryan Kastner & Dr. Curt Schurgers at Jacobs School of Engineering
  • Awarded Halıcıoğlu Data Science Institute Undergraduate Research Scholarship with Dr. Ryan Kastner
  • Working on low power microcontrollers to develop neck collars to monitor Panda’s behaviors using deep learning @KRG
  • Building tinyML solution for real-time animal sound inference on STM32, using MCUNET for CNN pruning & quantizing
  • Published a research paper to the NeurIPS climate change workshop and presented at the conference
• Research Assistant, Rady School of Management @ UC San Diego (Aug 2023 — June 2024)
  • Paid On-campus position leading technical development of a sleep study with over 30,000 participants across over 100 US college campuses through cross-platform app development using frameworks like React and Flutter
  • Conducting advanced NLP research for a national research project about PPP with several terabytes of projected data using Numpy, Pandas, TensorFlow, Pytorch and Scikit on the Narrows Cluster at San Diego Supercomputer Center (SDSC)
  • Integrated external APIs, including Fitbit's services, GCC & Firebase, for seamless sleep data retrieval and analysis
• Software Engineer Intern, doubleSlash Net-Business GmbH, Munich, Germany (Jul 2020 — Dec 2020)
  • Worked on my own backend project for internal IoT devices leveraging Dotnet, Microsoft Azure, C# and FFmpeg under the direct supervision of the Senior Developer and Microsoft MVP Ralf Richter
  • Assisted my team with client management and client products, creating several important features to in-house tools/products used by BMW and Porsche, such as the Navigation system, Light system and control system
• Head Of Software Development, TwoTronic GmbH, Meitingen, Germany (Feb 2020 — Dec 2020)
  • Worked in a leadership position, supervising the software development of the Vehicle Scanner 3.0
  • Developed an iOS application for Porsche by building a full-stack application for their internal App Store under the mentorship of Porsche’s Head of Logistics and Central Services, Sascha Drechsel
  • Led the development of the integration with the internal Mercedes Benz network API for low latency direct API calls and transfer of Big Data directly from the built in servers on over 20 vehicle scanners deployed across Germany
  • Led a team of 5 AI Engineers working on a custom Computer Vision damage detection algorithm by modifying and training the existing CNN (VGG-19 model) on a custom-built supercomputer with over 50TB of training data from vehicle scanners at Amazon warehouses and Mercedes Benz achieving over 90% accuracy

• PaintCalcArt, Los Angeles, CA – Full-Stack mobile app platform with over 5000 downloads and over 100 daily active users that can accurately detect and calculate the number of colors in pictures to return the amount of paint needed for bigger mural projects through custom Computer Vision ML algorithms using TensorFlow, Scikit & Pytorch integrated with Firebase Auth, Firestore, BigQuery, Flutter and GCC serverless functions.
• Growth Planner, Los Angeles, CA – Cross-platform mobile app with over 10,000 downloads and over 300 daily active users built with Flutter and designed with SOLID design principles in test-driven development for college and high school students to plan out every day in collaboration with Harvard student and famous Youtuber John Fish.

Languages: Python, Java, C#, C, C++, SQL, JavaScript, Matlab, Dart, Assembly, Swift, Objective-C, CUDA & OpenCL, Rust
Software & Tools: OpenCV, TensorFlow, Pytorch, Microsoft Azure, GCC, Docker, Scikit, AWS, Pandas, Git, Flutter, Numpy
Awards: Won 1st place in the computer vision challenge and 3rd place overall at a 24-hour hackathon hosted by Oerlikon by leveraging OpenCV over custom-trained AI, allowing real-time identification of manufacturing parts. Also constructed a full physical unit in the makerspace that would allow full automation with motors.