About this role
The Staff Researcher in AI Compute and Data Infrastructure conducts applied research and hands-on development for intelligent, efficient, resilient Hybrid AI systems. The role independently owns substantial research and development workstreams spanning AI algorithms, distributed computing, computer systems, and data infrastructure, with a focus on performance, scalability, reliability, and energy efficiency.
Responsibilities
- Build production-quality software, system components, services, APIs, diagnostic tools, and scalable data-processing pipelines. Characterize AI training, inference, and data-processing workloads through profiling, tracing, benchmarking, telemetry, logs, and hardware performance counters.
- Diagnose bottlenecks and reliability issues across GPUs and other accelerators, CPUs, memory, storage, networks, operating systems, runtimes, and AI frameworks. Develop hardware/software co-optimizations for utilization, scheduling, resource allocation, communication, data movement, storage access, and model execution.
- Optimize data ingestion, preprocessing, transformation, storage, retrieval, and delivery. Develop infrastructure diagnostics and resilience mechanisms for anomaly and regression detection, root-cause analysis, capacity forecasting, predictive maintenance, fault isolation, checkpointing, recovery, failover, and automated remediation.
- Apply machine learning, deep learning, time-series analysis, signal processing, causal inference, and knowledge graphs to workload modeling, performance prediction, operational decisions, and system health. Collaborate with hardware, systems, software, architecture, and product teams to bring research into Enterprise AI and Personal AI products; contribute patents, publications, reports, and reusable software, and mentor junior colleagues.
Minimum qualifications: A bachelor's degree in computer science, computer engineering, artificial intelligence, electrical engineering, applied mathematics, or a related field, or equivalent practical experience; three or more years of relevant experience in AI compute and data infrastructure, machine learning systems, distributed systems, data platforms, performance engineering, reliability engineering, or advanced software development. Strong programming skills in Python, C++, Java, Go, Rust, Scala, or a comparable language; ability to build and optimize reliable software; experience with profiling, telemetry, observability, performance diagnosis, or failure analysis; and expertise in at least two listed areas spanning ML, accelerator optimization, distributed AI, data processing, co-optimization, time-series methods, reliability, causal inference, or knowledge graphs. Strong analytical, experimental, communication, and collaboration skills are required.
Preferred qualifications
Experience with PyTorch, TensorFlow, JAX, CUDA, ROCm, Spark, Flink, Ray, Kafka, Kubernetes, or related technologies; cloud, edge, on-premises, or hybrid AI infrastructure; production delivery of research; or publications, patents, open-source contributions, or product impact.
This full-time hybrid role requires three onsite days per week in Morrisville, North Carolina. Candidates must be US citizens or US nationals; permanent residents and candidates requiring sponsorship cannot be considered.