Yuchen Tian
Adliswil, Zurich, Switzerland
Summary
🤩
Rockstar🎓
Top SchoolYuchen Tian is a curious, craftsmanship-driven software engineer with 13 years of experience building full-stack web and IoT platforms across finance, energy, and enterprise environments. He has shipped production services using Python, Flask, Angular/React, AWS (EC2, VPC, Cognito, S3) and Docker, and has experience designing testing and CI practices in Agile teams. Passionate about machine learning, Yuchen is the author of zi2zi, an open-source conditional GAN project for Chinese character style transfer, where he implemented core architecture improvements like conditional instance normalization and interpolation. His work blends practical backend and cloud engineering with ML research experiments, reflecting a hands-on approach from prototyping to production. Based in Zurich, he combines a Carnegie Mellon CS background with global delivery experience for large user bases and cross-border teams. Notably, he pairs legacy automation skills (VBA/OLAP) with modern ML and cloud practices, showing a rare breadth across tooling eras.
13 years of coding experience
3 years of employment as a software developer
Bachelor of Engineering, Computer Software Engineering, Bachelor of Engineering, Computer Software Engineering at Beihang UniversityMaster of Science (M.S.), Computer Science, Master of Science (M.S.), Computer Science at Carnegie Mellon University
Chinese, English
Github Skills (12)
neural-network10
neural-style-transfer10
machine-learning10
convolutional-neural-networks10
deep-learning10
tensorflow10
python10
generative-adversarial-network10
style-transfer10
image-processing10
model-optimization9
data-preprocessing9
Programming languages (5)
Github contributions (5)
kaonashi-tyc/Rewrite
Oct 2016 - Apr 2017
Neural Style Transfer For Chinese Characters
Role in this project:
ML Engineer
ML EngineerContributions:20 commits, 5 pushes, 2 comments in 5 months
Contributions summary:Yuchen primarily contributed to the development and improvement of a neural style transfer model for Chinese characters. Their work focused on refining the model architecture, including refactoring the network structure with different convolutional layer sizes, making the layer number configurable, and experimenting with leaky ReLU activation. The user also addressed data preprocessing and training aspects, adding data random shuffle and optimizing the training process. The contributions indicate a focus on improving model performance and flexibility.
neural-style-transfer
kaonashi-tyc/zi2zi
Feb 2017 - Aug 2019
Learning Chinese Character style with conditional GAN
Role in this project:ML Engineer
ML EngineerContributions:71 commits, 6 PRs, 15 pushes in 2 years 6 months
Contributions summary:Yuchen contributed to the development of a conditional GAN model for Chinese character style transfer. Their commits involve modifications to the `model.py` file, indicating work on the core architecture of the neural network, including encoder, decoder, and discriminator components. Furthermore, they added features such as conditional instance normalization and interpolation capabilities to enhance the model's performance.
chineseconditionalganconditional-gancharacter
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.