Senior Software Engineer, AI Frameworks

NVIDIA•Us, CA

About The Position

We are seeking a Senior Software Engineer to drive integration of the NVIDIA Grove project within Dynamo and across a set of leading open-source AI frameworks. In this role, you will develop production-grade software enabling Grove capabilities to be adopted, scaled, and operated smoothly. In this role, you will build production-grade software that enables seamless adoption, scaling, and operation of Grove capabilities across environments such as Dynamo, llm-d, Ray, PyTorch, and other emerging frameworks in the AI ecosystem. You will collaborate across engineering teams and the open-source community to deliver robust integrations, reference implementations, and developer-focused tooling.

Requirements

BS/MS/PhD in Computer Science, Electrical Engineering, or related field (or equivalent experience)
5+ years of proven experience in related field
Hands-on experience integrating with at least one major AI framework/runtime (e.g., PyTorch, Ray, Triton Inference Server ecosystem, distributed runtimes, model serving stacks).
Solid understanding of AI workloads: model development basics, training vs. inference tradeoffs, and performance considerations (throughput/latency, batching, memory).
Experience with distributed systems concepts (RPC, scheduling, fault tolerance, resource management).
Practical Kubernetes experience: deploying and operating services/jobs, Helm/Kustomize, operators/controllers (nice to have), and debugging clusters.
Familiarity with containers and cloud-native tooling (Docker, container registries, CI/CD pipelines).
Strong software engineering experience in Go, C++ and/or Python, with a track record of shipping reliable systems.
Strong interpersonal skills and ability to collaborate across teams and with open-source communities.
Exceptional collaboration, communication, and documentation habits.

Nice To Haves

Open-source contributions to Dynamo, PyTorch, Ray, llm-d, Kubernetes ecosystem, or related ML infrastructure projects.
Experience with large-scale model serving, distributed inference, or multi-tenant AI platforms.
Experience building SDKs/APIs or developer tooling that improves integration usability.
Knowledge of GPU performance profiling and optimization (Nsight tools or similar), and/or kernel-level performance tuning.
Experience with reproducibility, packaging, versioning, and compatibility testing across fast-moving dependencies.

Responsibilities

Design and implement end-to-end integrations of Grove with open-source AI frameworks (e.g., Dynamo, llm-d, Ray, PyTorch, and related ecosystem projects).
Build and maintain adapters, plugins, operators, and/or runtime components that enable Grove features to work smoothly across training and inference stacks.
Partner with framework owners to upstream changes, contribute patches, and ensure long-term maintainability of integrations.
Develop reference workflows, sample apps, and best-practice guides that accelerate adoption by users and partners.
Optimize performance, scalability, and reliability for distributed training/inference, including multi-node and multi-GPU environments.
Improve observability and operational readiness (metrics, logging, tracing, debugging tools) for Kubernetes-based deployments.
Participate in technical design reviews, define APIs/contracts, and ensure compatibility across versions of frameworks and dependencies.
Diagnose complex issues spanning containers, networking, scheduling, CUDA/GPU utilization, and framework runtime behavior.