I’m a Machine Learning Engineer and applied AI researcher focused on building end-to-end intelligent systems that translate research into real-world impact.

• Large Language Models (LLMs), Retrieval-Augmented Generation (RAG)
• Adaptive learning systems and AI-driven education platforms
• Multimodal deep learning (vision + time-series)
• Scalable ML pipelines, deployment, and performance optimization

• Machine Learning Engineer / Technical Lead at StudyRadar.ai

  • Architected and shipped an AI-powered exam preparation platform
  • Built RAG pipelines, real-time STT/TTS, and automated feedback systems
  • Designed scalable, cost-aware ML infrastructure on GCP

• Graduate Researcher (MSc Computer Science)

  • Designed BoRefAttnNet, a boundary-refined 3D U-Net for medical image segmentation
  • Developed DySTTM, a dynamic spatial-temporal transformer for multimodal learning
  • Paper accepted for peer-reviewed publication

• Selected as one of 10 Canadian researchers by the Digital Research Alliance of Canada
• Attended the 2025 International High-Performance Computing Summer School (IHPCSS) in Lisbon
• Training in GPU acceleration, parallel computing, and performance optimization

• Python, PyTorch, TensorFlow (Google TensorFlow Developer Certified)
• Transformers, RAG, FAISS, LangChain
• Docker, CI/CD, MLflow, GCP
• Research-to-production ML workflows

Most of my recent production work lives in private repositories due to IP and contractual constraints (e.g., StudyRadar.ai and consulting projects). Public repositories reflect research implementations, system prototypes, and experimental work. I am always happy to walk through architectures, design decisions, and trade-offs in detail.

• Email: imarhiagbeosasumwen@gmail.com
• LinkedIn: https://www.linkedin.com/in/osasumwen