Nica Liu
Senior Data Scientist
Chaoyang District, Beijing, China
Summary
🤩
Rockstar🎓
Top SchoolNica Liu is a Senior Data Scientist with eight years of experience combining business intelligence, product analytics, and applied machine learning at leading internet companies including Bytedance and Baidu. She pairs an MBA from Columbia Business School and earlier finance/accounting training with hands-on ML engineering—her open-source contributions include performance-focused work on mobile deep-learning inference (XiaoMi/mace) and data-quality metrics for federated learning (bytedance/fedlearner). At Sina Miaoche she led growth and product strategy across acquisition channels and national sales, demonstrating rare fluency across analytics, go-to-market execution, and business operations. Known for pragmatic model deployment and data validation work, she focuses on turning messy production data into robust, measurable ML inputs. Based in Chaoyang District, Beijing, she blends strong analytic rigor with product instincts to drive data-informed decisions at scale.
8 years of coding experience
6 years of employment as a software developer
Bachelor, Accounting & Financial Management, Bachelor, Accounting & Financial Management at Hiram CollegeMBA, Finance, MBA, Finance at Columbia University - Columbia Business SchoolBachelor-unfinished, Business Administration, Bachelor-unfinished, Business Administration at Beijing Foreign Studies University
Github Skills (14)
opencl10
machine-learning10
tensorflow10
data-validation10
performance-optimization10
python10
federated-learning10
data-pipelines9
data-engineering9
cprogramming-language9
data-pipeline9
c-language9
protobuf8
protobuffer8
Programming languages (5)
Github contributions (5)
bytedance/fedlearner
Jan 2021 - Jan 2023
A multi-party collaborative machine learning framework
Role in this project:
ML Engineer
ML EngineerContributions:27 reviews, 335 commits, 56 PRs in 1 year 11 months
Contributions summary:Nica implemented and integrated input data metric statistics for the fedlearner framework. They modified the data join CLI, raw data partitioner, and joiner implementations to include features for data validation, including a sample ratio and optional fields. The contributions involved modifications to protobuf definitions and the introduction of a `MetricStats` class to compute and emit metrics for the input data, alongside date-based filtering. These changes enhanced data quality and provided insights into the input data characteristics within the federated learning context.
pythonpartydata-sciencedeep-learningmachine-learning
XiaoMi/mace
Sep 2017 - Jun 2019
MACE is a deep learning inference framework optimized for mobile heterogeneous computing platforms.
Role in this project:ML Engineer
ML EngineerContributions:3 releases, 710 commits, 15 PRs in 1 year 9 months
Contributions summary:Nica's commits primarily focus on enhancing the MACE deep learning inference framework, specifically regarding half-type const tensor support and optimizing OpenCL kernel implementations. They introduced modifications to utilize half-type tensors within the framework, including adjustments to serialization processes and support in various ops test files. Furthermore, the user refactored NEON and OpenCL implementations, suggesting a focus on performance optimization and supporting efficient execution of deep learning models on mobile platforms.
neonpytorchheterogeneous-computingdeep-learning-inferenceheterogeneous
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.