Descripción de la oferta
Die PositionAs a User Enablement Engineer within the Accelerated Compute Engineering (ACE) team, you will be responsible for overseeing the deployment, optimization, and day‑to‑day operation of user‑facing software, application frameworks, and research environments across our High‑Performance Computing (HPC) and AI Factory platforms. Your mission is to maximize the productivity of Roche’s scientific community by providing frictionless access to cutting‑edge computational tools.With the introduction of our industry‑leading NVIDIA AI Factory infrastructure, you will play a pivotal role in productizing and supporting advanced software suites like NVIDIA AI Enterprise. By ensuring that complex AI/ML frameworks, large language model (LLM) tools, and specialized simulation environments are perfectly tuned for our clusters, you will directly accelerate the time‑to‑value for researchers training complex AI models and executing large‑scale computational science workloads.Description of the areaHosting and Infrastructure (HI) provides mission‑critical on‑premise infrastructure, cloud hosting, connectivity, and technology products that enable all functions at every Roche site to develop, innovate, connect, and deliver compliant digital products across the Roche Enterprise.The Value Streams – Accelerated Compute Engineering (ACE) Team is focused on driving both customer success and platform success by acting as a centre of excellence and delivery for the High Performance Compute and AI Infrastructure supporting AI and HPC use cases across Roche. This team facilitates seamless onboarding and adoption for business vertical customers needing accelerated compute—helping those infrastructure consumers with needs optimised for high availability, seamless data transfer, flexibility, speed, and the rapidly changing needs of AI—helping achieve rapid time‑to‑value.Job ResponsibilitiesUser‑Facing Platform & Environment ManagementDeploy, support, and optimise user‑facing environment portals and platforms, including Roche ROCs and SCOOP environments for HPC.Own the lifecycle and availability of standard software stacks, modules, and containerised application environments.Collaborate closely with infrastructure engineers to ensure user‑facing portals interface seamlessly with core scheduling and data transfer layers.AI Factory & Software Stack ProductisationArchitect and manage the deployment of the NVIDIA AI Enterprise software suite, ensuring proper integration and licence optimisation.Provide and support cutting‑edge AI pipelines and development platforms, including NVIDIA NeMo for generative AI/LLMs and NVIDIA Omniverse for advanced industrial simulations and digital twins.Build, benchmark, and maintain highly optimised container builds for standard AI/ML frameworks (such as PyTorch and TensorFlow), incorporating deep learning inference servers like Dynamo‑Triton to maximise GPU utilisation.Scientific Consultation & Technical EnablementIn partnership with the Value Engineering and Outcomes team, act as the technical entry point and consultant for research teams, assisting them in porting their code, algorithms, and models onto the ACE platforms.Develop, curate, and maintain developer documentation, example code repositories, and user guides to enable self‑service onboarding.Diagnose and resolve complex software‑layer bottlenecks, dependency conflicts, and compilation issues spanning scientific software packages, Python virtual environments, and GPU‑accelerated libraries (e.g. CUDA, cuDNN).QualificationsEducation / Experience:Bachelor’s or an advanced degree in Computer Science, Data Science, Bioinformatics, Computational Chemistry, or a similar technical discipline.5+ years of experience in an engineering role focused on scientific software compilation, user enablement, or DevOps for advanced research compute environments.Deep proficiency navigating and managing user applications within Enterprise Linux (RHEL/Ubuntu) multi‑tenant systems.Technical & Business Skills:AI/ML & Inference Tooling: Experience configuring and optimising NVIDIA AI Enterprise applications, specifically NeMo, Omniverse, and Triton Inference Server architectures.Framework Mastery: Deep understanding of modern AI/ML frameworks (PyTorch, Jax, TensorFlow) and their performance tuning on modern GPU architectures.Environment Provisioning: Proficiency with environment management tools (such as Lmod/Environment Modules, Conda) and container technologies (Singularity/Apptainer, Docker).Scripting: Strong scripting capabilities (Python, Bash) to automate user environment configurations and application testing pipelines.Enforce a strict configuration‑as‑code and infrastructure‑as‑code mindset, replacing manual interventions with repeatable automation scripts.Leadership & Mindset:Lean & Agile Mindset: Highly focused on driving self‑service capabilities and automating application delivery to scale support.Customer Champion: A genuine passion for working with researchers, understanding their technical pain points, and translating complex infrastructure into accessible services.Intelligence & Curiosity: A strong passion for innovation and staying current with the rapidly changing open‑source and commercial AI/ML software landscape.Roche ist ein Arbeitgeber, der die Chancengleichheit fördert.
#J-18808-Ljbffr
