Mingsheng Hong
Head VP Of AI at Ironclad
San Francisco Bay Area United States
Summary
🤩
Rockstar🎓
Top SchoolMingsheng Hong is a seasoned AI and data engineering leader with 11 years of experience building systems that move ideas from prototype to product-market fit and commercial scale. He has led ML runtime and data infrastructure teams at Google, founded and scaled Bluesky to a $1.5M+ ARR exit to Microsoft, and now leads AI product strategy and execution as Head/VP of AI at Ironclad. His work spans deep infrastructure (TensorFlow/JAX runtimes, Swift for TensorFlow contributions) through customer-facing AI-native platforms where data quality, evaluation, and reliability drive real business impact. Comfortable as both a hands-on engineer and executive, he blends research-grade systems engineering with product and go-to-market rigor. Unusually, he pairs low-level runtime contributions (e.g., TensorFlow runtime tests and RL model work) with founding-stage GTM experience, enabling technical decisions that directly improve adoption and monetization.
11 years of coding experience
13 years of employment as a software developer
BSc Computer Science, BSc Computer Science at Fudan UniversityPh.D. Computer Science, Ph.D. Computer Science at Cornell University
English, Chinese, Japanese
Github Skills (14)
asynchronous10
unit-testing10
machine-learning10
swift10
c-language10
tensorflow10
cprogramming-language10
reinforcement-learning10
async10
python8
optimizers7
optim7
adam7
optimizer7
Programming languages (3)
Github contributions (5)
tensorflow/swift-models
May 2018 - Mar 2019
Models and examples built with Swift for TensorFlow
Role in this project:
ML Engineer
ML EngineerContributions:7 commits, 15 PRs, 20 pushes in 9 months
Contributions summary:Mingsheng contributed to the development of machine learning models within the repository, focusing on models using Swift for TensorFlow. Their work includes fixing bugs in an MNIST model, specifically improving loss calculations and addressing compiler issues. Additionally, the user implemented a reinforcement learning model for the CartPole problem, optimizing performance by switching optimizers and addressing feedback. The user also contributed to solving the FrozenLake RL problem using Q-learning and Python integration.
swifttensorflowswift-for-tensorflow
tensorflow/runtime
Apr 2020 - Apr 2021
A performant and modular runtime for TensorFlow
Role in this project:Back-end Developer
Back-end DeveloperContributions:11 commits, 21 comments in 11 months
Contributions summary:Mingsheng primarily contributed to the core runtime environment of TensorFlow, adding unit tests and expanding the system's capabilities for composite operations. Their work involved modifying existing components to support new functionalities, evidenced by changes in core runtime files like `core_runtime.cc` and `core_runtime_op.cc`. Furthermore, the user refactored the test library and added debugging support for the `AsyncValue` class, indicating their efforts to improve the project's testability and debuggability. They also implemented changes to support native composite ops.
runtimeperformantmodulartensorflow
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.