Hi, I'm Yutong—an inquisitive University of Michigan Data Science major (May 2026) who turns complex systems into approachable tools. Recently, I shipped optimization pipelines as a Software Engineering Intern at Shenzhen Taikejia Technology and prototyped playful generative-art experiences with cross-disciplinary design partners.
My sweet spot lives at the intersection of machine learning, optimization, and thoughtful UX. Whether I'm profiling a branch-and-bound solver, tuning a hybrid CNN–ViT model, or curating travel photo essays, I chase outcomes that are rigorous, kind, and creative.
Research sprint coordination, cross-functional scheduling, requirements tracking, risk mitigation.
Python, C++, Java, JavaScript, R, SQL, LaTeX.
PyTorch, NumPy, Pandas, MATLAB, STATA, Git, Linux/Unix workflows.
English, Mandarin Chinese, Cantonese.
Exploring light, color, and quiet moments through street and travel photography.
Curating playlists and discovering indie artists that inspire new creative moods.
Staying active with fast-paced rallies and the teamwork that comes with doubles play.
Balancing busy weeks with strength training and mindful movement.
Giving back through community initiatives that support education and equity.
Each build starts as a playful sketch in my notebooks—then becomes a living system I can ship, measure, and iterate on. Here are a few capsules from that design-to-deployment arc.
A cross-platform branch-and-bound solver with pybind11 bindings so analysts can prototype models in Python, then burst into C++ performance when it’s time to ship.
React + Flask experience with real-time feeds, infinite scroll, and async rendering—designed so stories stay on screen while data flows from modular REST endpoints on AWS.
Hybrid CNN + ViT pipeline trained on 8,867 labeled photos, reaching 93% validation accuracy with focal loss tuning, mixup augmentation, and model explainability overlays.
A visual demo board that captured motion studies, color stories, and UI beats for a generative art experience before translating the system to engineering specs.
R Shiny environment for UN demographic planning with scenario toggles, quick-glance summaries, and narrative annotations for policy partners.
A collection of my favorite photographs and moments.
I log the playbook behind the site—strategy, design, and launch routines that help me stay intentional while shipping quickly.
How I prioritize the essential pages—homepage, about, portfolio, resume, and contact—before layering on extras like testimonials or services.
Thoughts on pairing warm sunrise gradients with twilight palettes, structuring navigation, and keeping mobile-first layouts feeling immersive.
A launch checklist that covers social integrations, SEO foundations, testing loops, and the habit of updating before things feel “perfect”.
Email: yutonglvv@gmail.com
Email: yutonglv@umich.edu
Email: yutonglv@outlook.com
GitHub: https://github.com/lytreallynb
Portfolio: https://lytreallynb.github.io
LinkedIn: https://www.linkedin.com/in/yutong-lyu
CV: Download PDF
Languages: Mandarin Chinese, Cantonese, English
Ask me anything about Yutong!