Rui Chen

Education

Infra: Proficient in Kubernetes, Docker, familiar with mainstream CI pipelines like GitLab/GitHub, experienced with various services on major cloud platforms such as AWS and Alibaba Cloud
Development: Expert in Python along with its standard library source code, familiar with most common third-party libraries; highly skilled in Linux environments and commands, capable of efficiently diagnosing and solving complex problems
Computer Science Fundamentals: Proficient in algorithms and data structures; research experience in NLP and CV
Passionate about open source, love to explore, always curious, often contribute to open-source projects in spare time

Responsible for the design and implementation of cloud infrastructure for RSD ingestion and Route Matcher services
In charge of China-specific development tasks for the above services
Responsible for deploying and maintaining LLM (vllm) in compliance room
Designed, developed, and deployed the architecture for taskXQ (compliance cloud offline environment task execution platform)

Developed the web platform using SpringBoot and MySQL with RESTful APIs and SSO login via OAuth 2
Created APIs for the user, order, and device management systems, handling CRUD operations with role-based authentication

Parsed ADCU C struct data, developed visualization layers, and integrated into the map system based on MVT
Debugged corner cases like memory leaks and prolonged non-responsiveness in production, solving business issues
Led the migration and maintenance from Nomad to Kubernetes, enhancing the maintainability of service clusters
Optimized deployment clusters based on service load curves, reducing costs by 60% while keeping response performance

Designed the business architecture of taskXQ, developed based on Nomad + Docker Swarm for on-demand resource allocation and execution of any tasks on the compliance cloud, providing the team with task automation for running, monitoring, and notifications
Task definition schema and interface are highly extensible, allowing seamless migration to a modern K8s + Karpenter architecture

Implemented automated execution of sign-off evaluations and exports based on taskXQ
Optimized business code logic, reducing average response time for data post-processing by 40%
Led deployment and maintenance migration to Kubernetes

Deployed quantized large language model (Qwen3-32B-AWQ) for internal use based on vllm + open-webui

Fixed the issue where models based on openrouter/claude couldn’t correctly use cache_control. PR: !8692
Resolved the problem of fn_call failures caused by the introduction of the browser tool. PR: !8695
Actively participated in upstream and downstream dependencies of open-source projects, engaged in community discussions, and contributed code

Developed using ADK + FastAPI, deployed on Alibaba Cloud using K8s with ALB Ingress
Implemented automated unit/integration testing, automated builds for multiple environments, and automated independent deployment for multiple development branches using GitHub Action pipelines