Senior RDMA Performance Engineer for AI Data Centers

Codilime; Polska; Hybrid; PLN 240,000 - 320,000; Full time

Job summary

Codilime is seeking an RDMA Performance engineer to join our project team focusing on AI networking environments. You will configure and troubleshoot NVIDIA/Mellanox Spectrum switches and ConnectX NICs, tuning congestion control for AI workloads and optimizing throughput and latency in data centers.

You’ll work with NCCL, RDMA, and QoS while automating tasks with AI agents and documenting your designs and benchmarking results for internal and client reviews.

Qualifications

• 7+ years of professional experience in network engineering, with 3+ years in data center design.
• Hands-on experience with BGP, EVPN/VXLAN, LACP and ECMP.
• Deep knowledge of AI-dedicated data center designs (ROD/RUD) and practical deployment.
• Proven experience configuring, managing and troubleshooting NVIDIA/Mellanox Spectrum switches and ConnectX NICs.
• Deep understanding of RoCEv2, PFC, and ECN for AI data centers.
• Ability to tune switch buffers and implement QoS policies.
• Experience benchmarking RDMA-enabled fabrics and NCCL.
• Ability to leverage AI agents/tools to automate tasks and testing.
• Strong technical documentation and professional English communication.

Responsibilities

• Configure, troubleshoot and maintain NVIDIA/Mellanox Spectrum switches and ConnectX NICs.
• Implement and tune congestion control for AI workloads (RoCEv2, PFC, ECN).
• Tune network environments for maximum throughput and minimal latency.
• Benchmark and validate AI-ready data center fabrics using RDMA and NCCL.
• Automate tasks using AI agents and testing tools; analyze telemetry.
• Document network designs, configurations and benchmarking results.

Skills

Network engineering

Data center design

BGP

EVPN/VXLAN

LACP

ECMP

QoS tuning

RDMA

NCCL

English communication

Tools

NVIDIA/Mellanox Spectrum switches

ConnectX NICs

pyATS

NAPALM

Batfish

Nornir

NetBox

CI/CD pipelines

Job description

Codilime is seeking an RDMA Performance engineer to join our project team focusing on AI networking environments. You will configure and troubleshoot NVIDIA/Mellanox Spectrum switches and ConnectX NICs, tuning congestion control for AI workloads and optimizing throughput and latency in data centers.

You’ll work with NCCL, RDMA, and QoS while automating tasks with AI agents and documenting your designs and benchmarking results for internal and client reviews.