Senior HPC and MLOps Engineer

In this role, you will play a key part in the development, optimization, and operation of the Institute’s AI-optimized HPC cluster (the AI Foundry), which integrates 68 Nvidia’s Hopper and 80 Blackwell AI accelerators plus 5PB of fast storage.

You will contribute to the design and implementation of MLOps workflows and AI development pipelines, supporting our researchers in deploying innovative solutions across the AI lifecycle. You will work closely with other R&D units to optimize infrastructure, troubleshoot complex issues related to model training and deployment, and help ensure the reliability, scalability, and performance of our AI systems.

Your key responsibilities:

• Technical Leadership: Contribute to the adoption and implementation of state-of-the-art AI and MLOps technologies, frameworks, and best practices. Collaborate in the deployment and maintenance of secure, scalable, and high-performance infrastructure for continuous model delivery and lifecycle management.
• Engineering Excellence: Develop and maintain reusable frameworks, libraries, and APIs tailored to industry needs. Support rapid prototyping and the operationalization of AI models.
• Collaboration: Work alongside researchers, engineers, and partners to deliver integrated AI solutions, troubleshoot issues, and ensure efficient workflows.
• Continuous Improvement: Monitor advancements in AI and HPC, suggest new tools and techniques, and support a culture of technical excellence and innovation within the team.
• Project Contribution: Support the execution of complex projects, ensuring technical quality, reliability, and alignment with project goals and deadlines.

What We Offer

• A unique opportunity to shape the future of AI research and industrial innovation in Italy and Europe
• Access to a state-of-the-art AI-optimized HPC facility
• Collaboration with top-tier researchers and engineers
• A stimulating and rewarding environment within a leading AI institute
• A competitive salary with significant performance-based incentives

Desired Qualifications

We are seeking candidates who possess:

• MSc in Engineering, Computer Science, Physics or other relevant fields (PhD is a plus)
• At least 5 years’ experience in MLOps, HPC, or related roles
• 10+ years of relevant experience in coding and systems development
• Experience with PyTorch
• Experience with distributed computing frameworks such as the message passing interface (MPI)
• Strong background in systems architecture and design, spanning Data, AI/ML, Core Infrastructure, and Security Engineering
• Experience in both cloud and on-prem or colocation hosting in Tier 2 data centers
• Extensive scripting experience
• Hands-on familiarity with Nvidia CUDA and custom AI/ML libraries
• Experience with systems deploying Nvidia Hopper GPUs (experience with Blackwell is a plus)
• Deep knowledge of Linux system optimization and administration
• Understanding of Data Engineering, Data Governance, Data Infrastructure, and AI/ML platforms
• Certifications, such as NCP‑AII (AI Infrastructure), NCP‑AIO (AI Operations) from the NVIDIA Certified Professional infrastructure, and other certified training specific to Nvidia Hopper and Blackwell are considered a strong plus.

Key Success Metrics

You will be evaluated based on:

1. Technical Contribution: Delivering high-quality, scalable, and efficient solutions aligned with AI4I’s goals.
2. Impact: Contributing to the design, deployment, and continuous improvement of AI engineering and MLOps infrastructure, leading to measurable increases in system efficiency, reliability, and speed of model delivery.
3. Collaboration: Supporting team and project goals through effective technical collaboration, troubleshooting, and knowledge sharing.

How to Apply:

Please submit to: jobs@ai4i.it

1. CV
2. Motivation letter (max. 1,500 words)
3. Three professional references (contacts).
4. Optional: Patents, publications, git, portfolio links.

Diversity & Equal Opportunity

AI4I is committed to building diverse, inclusive teams. Applications from underrepresented groups are strongly encouraged.