• Research & Benchmarking: Analyze current ML models (TF-IDF/Random Forest) in comparison to state-of-the-art Deep Learning and LLM architectures
  • Multimodal Development: Design and prototype models that process text as well as technical drawings and metadata
  • Performance Optimization: Define metrics to increase accuracy in identifying export control sensitivities
  • Prototyping: Implement Proof-of-Concepts (PoC) and integrate them into the existing AI tool
  • Cross-functional Collaboration: Coordinate with Data Scientists, Engineers, and Legal Experts to ensure regulatory compliance.