Senior AI Research Engineer (GPU Systems & Kernels)

Location: Paris, France; hybrid work

Company

Datadog provides an observability and security platform that unifies applications, infrastructure, data, models, and security, using AI to detect and resolve issues.

What you will do

• Write, tune, and debug GPU kernels in CUDA, Triton, or similar technologies for LLM training and inference.
• Optimize workloads for current NVIDIA GPUs, including memory hierarchy, tensor cores, FP8, and FP4.
• Reduce distributed communication costs across all-to-all and all-reduce workloads, including communication and compute overlap across NVLink and multiple nodes.
• Profile training and rollout workloads to identify compute, memory, and communication bottlenecks and quantify optimization headroom.
• Build benchmarks, profiling tools, and regression tests, and collaborate with Research Scientists to implement new model architectures efficiently.
• Explore low-precision training and serving, speculative decoding, and fused or persistent kernels.

Requirements

• Depth in at least one of GPU kernel development and optimization, low-latency communication and collective operations, or large-scale training infrastructure.
• Ability to reason about memory, bandwidth, and interconnect topology, identify bottlenecks, and define how to validate them.
• Proficiency in a systems language such as C++, Rust, or CUDA C, and in Python.
• Strong debugging and profiling skills, including tools such as Nsight.
• Ability to explain design and performance trade-offs clearly to technical and non-technical audiences.

Nice to have

• Experience with Mixture-of-Experts models, expert parallelism, or sparse model inference.
• Familiarity with NCCL, NVSHMEM, and compute/communication overlap.
• Experience with Triton, CUTLASS, or similar kernel frameworks.
• Familiarity with Megatron-LM, TorchTitan, DeepSpeed, or other distributed training frameworks.
• Experience building GPU-intensive systems for graphics, rendering, simulation, or scientific computing.
• Contributions to open-source ML or systems infrastructure projects.

Culture & Benefits

• Work on high-risk, high-reward AI research problems grounded in cloud observability and security.
• Collaborate with Research Scientists on infrastructure for trained agents and large language models.
• Datadog supports applicants from diverse backgrounds and recognizes that candidates may not meet every qualification on day one.
• Equal employment opportunity is provided regardless of protected personal characteristics, consistent with legal requirements.