Milad Sikaroudi
Senior Machine Learning Engineer @Roche
Signup · Get unlimited contacts
WORK HISTORY
Senior Machine Learning Engineer @Roche
EDUCATION
Tehran University of Medical Sciences
Master's Degree
University of Waterloo
Doctor of Philosophy - PhD
ABOUT MILAD SIKAROUDI
Highly accomplished Machine Learning and Software Engineer with a Ph.D. in Systems Design Engineering, specializing in AI systems, deep learning, MLOps, and large-scale image processing. Proven expertise across the full prototype-to-production lifecycle, including shipping AI containers/products, deploying them in enterprise environments, and delivering solutions to major corporate partners. As an ML Engineer at Roche, I design, develop, and train/deploy AI products in digital pathology, leading training models, containerized inference pipelines, continuous versioning, and automated deployment workflows. Skilled in GitLab/GitHub CI/CD, Python–C++ performance optimization, W&B model registries and pipelines, and building integrated end-to-end AI ecosystems across complex enterprise infrastructures.A critical thinker with a strong bias for action, I proactively identify inefficiencies, automate repetitive processes before they become bottlenecks, and support stakeholder decision-making through clear engineering insight. Known as a collaborative and reliable team member, effective across multidisciplinary groups—from engineering to product and business. Driven by impact, ownership, and continuous improvement.Previously contributed to the development of a real-time surgical navigation system, leveraging expertise in design patterns, agile/scrum methodologies, clean architecture, modern C++, build systems, SQL, and GLSL.Technical proficiencies include Python, C/C++, JavaScript/TypeScript, SQL; Git; Azure DevOps; TDD; CI/CD; and shell/build systems. Fluent with PyTorch/Lightning, Triton Inference Server, ONNX runtime, TensorFlow, scikit-learn, and experienced with AWS/GCP, containerized deployment, W&B, FastAPI, Docker, Triton custom backends, and CUDA.As a researcher, strong background in domain generalization, few-shot/meta-learning, federated learning, multimodal learning, vision transformers, and weakly/self/unsupervised learning.I published in MICCAI, Nature, Transactions, and ICCV. Passionate about applying AI and computer vision to transform healthcare and improve patient outcomes, and continuously seeking opportunities to contribute to impactful, cutting-edge projects.
This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.