GPU engineer - REMOTE - UN-agency-client - EN-speaking

We are looking for one "remote GPU Engineer - GPU Workload Optimization expert" (Freelance) with NVIDIA Certification (preferred) to start ASAP an contract until years end (extensions possible) for an international Org' - client in ROME, where 100% REMOTE work is possible. Possibility of extension based on performance and budget availability. English speaking environment.

• Our client is an international organisation in Rome and a great reference in any CV!
• On-Site in Rome would be great, but up to 100% Off-site/Off-shore is possible.
• A competitive rate will increase chances to be selected.

ROLE

• As a GPU Engineer, you will design, implement, and optimize GPU-accelerated workloads to support AI and data-driven use cases.
• You will benchmark, profile, and tune GPU performance, automate GPU resource provisioning, and contribute to Infrastructure as Code (IaC) best practices.
• The role also involves developing reference implementations, enabling self-service deployment of LLMs, and knowledge transfer to ICT staff.

TASKS

• Benchmark, profile, and tune GPU workloads; recommend sharing techniques (MIG, vGPU, MPS).
• Optimize applications via GPU parallelization and framework-level improvements.
• Evaluate deployment methods for GPU-accelerated serving frameworks (eg, NVIDIA Triton, TensorRT, ONNX Runtime).
• Automate GPU resource provisioning and monitoring with IaC tools (Terraform, Helm, Kubernetes).
• Establish GPU dashboards and proactive health/performance monitoring.
• Support deployment of LLMs on GPU resources.
• Document workflows and provide training/knowledge transfer.

SKILLS

• >2 years of hands-on GPU engineering or cloud-based GPU workload optimization.
• NVIDIA Certification (preferred).
• GPU-accelerated development (CUDA, TensorRT, PyTorch, TensorFlow, ONNX, etc.).
• Benchmarking/profiling expertise (Nsight, nvprof, etc.).
• IaC experience (Terraform, Helm) and CI/CD pipeline development (GitHub Actions preferred).
• Kubernetes GPU scheduling and workload deployment.
• Monitoring via Prometheus, Grafana, DCGM, or similar.
• Strong Scripting skills (Python, Bash, PowerShell).
• Excellent troubleshooting, problem-solving, and teamwork abilities.

AWARD & OTHERS

• The applicants attention is drawn to the important role that the curriculum vitae plays in the evaluation.
• Curriculum vitae shall illustrate the specific skills relevant to this request.
• We would like to receive CVs of suitable candidates together with pricing quotations, based on a daily net rate.
• Please apply when you are available immediately or on short notice (and Full-time).
• Please respect, that I am NOT working with agencies/companies.

PROVIDER

infom consulting is an owner-managed business and consulting firm in Germany. The company supports large corporations and larger SMEs across Europe. Our IT experts are realising projects for the European Institutions, United Nations agencies, International Organisations and multinational companies across the EU.

TO DO - QUESTIONS

1. INTRO': Can you please provide a 3-4 sentences summary how you fit It's a kind of Intro' to be used for your presentation to the UN-client, which will help a lot.
2. RATE(S): Can you tell me your daily rate (all in) in Euro, please? 100% OFF-site work will be possible.
3. CV: Can you send your CV in DOC-format highlighting the specific skills relevant to this request, please?
4. Can you tell me your availability (or notice period), please? We need one single key-consultant, who needs to be available full-time, 8 hrs per working day, Mon-Fri, 40 hrs/week, to work with team.