Yuqi Li
Software Engineer at Google
Sunnyvale, California, United States
Summary
🤩
Rockstar🎓
Top SchoolYuqi Li is a software engineer with 11 years of experience specializing in on-device machine learning and model deployment, currently working on MediaPipe and TensorFlow Lite at Google in Sunnyvale. He combines backend and ML engineering skills demonstrated by significant open-source contributions to flagship projects like Apache MXNet (R bindings and RNN work), TensorFlow models/tflite-support (object detection, TFLite export and metadata), and MediaPipe metadata tooling. His work spans from model architecture and optimization (quantization, full-integer TFLite export) to practical deployment concerns such as metadata, tensor mapping, and mobile inference. A Peking University alumnus with multiple internships at major tech firms, he brings both research-grounded understanding and production-focused pragmatism to edge ML problems. Notably, he has deep experience adapting complex NLP and vision models for resource-constrained devices, reflecting a knack for making heavyweight models practical on mobile and IoT platforms.
10 years of coding experience
1 year of employment as a software developer
Bachelor's degree, Computer Science, Bachelor's degree, Computer Science at Peking University
Github Skills (38)
application-framework10
gru10
python10
image-classification10
r10
app-framework10
mxnet10
machine-learning10
rnn-model10
metadata10
n10
deeplearning-ai10
lstm10
tensorflow-lite10
deep-learning10
Programming languages (5)
Github contributions (5)
tensorflow/examples
Jan 2020 - Oct 2022
TensorFlow examples
Role in this project:
ML Engineer
ML EngineerContributions:217 commits, 2 branches in 2 years 10 months
Contributions summary:Yuqi refactored the image classifier model, changing how the model specification is handled by replacing the hardcoded model name with the model spec, which provides greater flexibility for model selection. The commit replaced the model name with model specification for increased versatility, improved the model by incorporating image data augmentation in preprocessing and added the full integer quantization for tflite. These changes involved modifications to the model's architecture and exporting process.
tensorflow
google-ai-edge/mediapipe
Oct 2022 - Jan 2023
Cross-platform, customizable ML solutions for live and streaming media.
Role in this project:ML Engineer
ML EngineerContributions:15 commits in 3 months
Contributions summary:Yuqi contributed to the migration and implementation of base metadata functionality within the MediaPipe framework. Their work involved transferring classes such as MetadataPopulator and MetadataDisplayer into MediaPipe and creating new metadata functionality. The commits show code modifications within the metadata.py file related to TensorFlow Lite metadata tools and the addition of score calibration file. These changes highlight the user's focus on improving and expanding metadata capabilities within the MediaPipe project.
neumorphismc-plus-plusvideotensorflowgraph-based
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.