Developed and delivered multi-node inference support for the vllm-project/aibrix repository, enabling scalable distributed LLM workloads using vLLM with the Qwen2.5-Coder-7B-Instruct model. The work involved designing and implementing head and worker node configurations, integrating an aibrix runtime sidecar, and defining service definitions and HTTP routing to support distributed inference. Leveraged Kubernetes and YAML to orchestrate resources, container commands, and networking, ensuring deployment readiness and reliability. Addressed integration gaps by providing a complete RayClusterFleet example, which streamlined onboarding and improved scalability for machine learning operations. Demonstrated depth in distributed systems, container orchestration, and model integration workflows.