Nan Wang
Global, China
Summary
🤩
RockstarNan Wang is a founder and co-founder & CTO at Jina AI with 11 years of experience building cloud-native, multimodal AI infrastructure from a global base in China. He combines founder-level product sense with hands-on backend engineering, contributing deep work on transformer encoders, model loading, masks, and driver/CompoundExecutor bug fixes in the flagship jina-ai/serve repository. He also implements end-to-end examples—adding indexing and query features and migrating demos to newer framework versions—demonstrating an eye for developer experience and production readiness. His contributions span code, tests, and documentation, signalling a pragmatic engineer who ships robust ML infrastructure while keeping the developer journey smooth.
12 years of coding experience
Github Skills (16)
transformer-models10
encode10
jina10
python10
semantic-search10
encoder10
tensorflow9
nlp9
deep-learning9
tensorflow29
machine-learning8
deep-learning-ai7
dockers7
docker7
fastapi7
Programming languages (8)
Github contributions (5)
jina-ai/examples
Mar 2020 - Sep 2021
Jina examples and demos to help you get started
Role in this project:
Back-end Developer
Back-end DeveloperContributions:164 reviews, 203 commits, 279 PRs in 1 year 6 months
Contributions summary:Nan primarily contributed to the development of the "urbandict-search" example. Their work focused on implementing features, including adding indexing and query functionality. They made changes to key files like `customized_executors.py` and `query.py`, which suggests interaction with Jina's executor framework and core application logic. Furthermore, the user adapted the project to newer versions of the Jina framework and refactored code.
nlppythononboardingget-starteddeep-learning
jina-ai/serve
Feb 2020 - Dec 2022
☁️ Build multimodal AI applications with cloud-native stack
Role in this project:Back-end Developer
Back-end DeveloperContributions:5 releases, 530 reviews, 469 commits in 2 years 10 months
Contributions summary:Nan's contributions focused on adding and improving the encoder functionality within the "serve" repository, which builds multimodal AI applications. They addressed bugs in drivers related to CompoundExecutor and implemented transformer encoders, including related test cases and documentation. Their work involved refactoring codes and fixing issues in masks and model loading, suggesting a strong understanding of transformer-based models.
opentelemetryllm-inferencecreative-aigrpcproduction
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.