Are you an AI Platform Engineer expert in the platform that turns a researcher's idea into a running job on a cluster?
Join Reply
WHO WE ARE
Reply specialises in the design and implementation of solutions based on new communication channels and digital media. As a network of highly specialised companies, Reply supports major industrial groups in the telecom and media; industry and services; banking and insurance and public sectors in defining and developing business models enabled by the new paradigms of AI, cloud computing, digital media and the internet of things. Reply's services include: consulting, system integration and digital services.
WHAT WILL YOU DO?
Core activities. You will keep the core of the Model Factory platform standing, and you will make it grow. You will orchestrate GPU workloads and distributed training that hold up under load, build versioning and governance across data, models and evaluation runs so results are still reproducible months later, and shorten the distance between a researcher's idea and a running job on the cluster. Every model we train,
distil or optimise — for our own portfolio and for client engagements — runs on what you build.
Tech & Tools Stack. You will work SkyPilot, Dagster, NVIDIA NeMo, Kubernetes, MLflow and Unity Catalog, the core of the platform. Around them: the NVIDIA stack (CUDA, NCCL), job scheduling across national (Italian) and European HPC clusters and cloud (such as Scaleway or Nebius), infrastructure as code and GitOps CI/CD, vLLM for inference serving, and Prometheus and Grafana observability that tells us what a run cost and why it failed.
Team work. You will work alongside the AI research engineers who train the models, the model design engineers who specify them, the project managers who commit the dates, and cloud and security specialists — plus client platform teams when the environment lands on their side. Your first customers are internal: if the platform is slow, opaque or fragile, they feel it lo
📌 AI Platform Engineer (Torino)
🏢 Reply
📍 Torino