Senior System Software Engineer – Accelerated Microservices and Libraries

We are looking for a senior software engineer to contribute to building accelerated microservices and libraries for our harnesses and agents.  In this role, you will be part of an engineering team designing and developing microservices and libraries and optimizing these for relevant metrics. 

What you will be doing: 

• Contribute to design and development of accelerated microservices and libraries 

• Collaborate with peer engineering and Product Management teams to understand requirements and translate these into deliverables across milestones. 

• Design, develop, and maintain real-time vision-language model inference microservices, from video ingestion and preprocessing through model serving and downstream integration. 

• Develop automated functional, accuracy, and performance regression tests, and own component readiness through integration, release, and post-release debugging. 

• Profile and optimize end-to-end latency, throughput, GPU memory usage, and concurrent user capacity across platforms. 

What we need to see: 

• Masters or Bachelors degree in Computer Science, or related engineering field  

• 8+ years of overall experience in software development , a strong technical foundation in software design principles  and strong programming skills in C/C++ and Python. 

• Hands-on experience building microservices using containers, REST APIs, and asynchronous messaging. 

• Hands-on experience integrating and serving computer vision and vision-language models in production. 

• Understanding of real-time video processing and the tradeoffs among latency, throughput, and resource utilization and strong debugging skills. 

• Experience evaluating model quality and diagnosing accuracy or performance regressions across model, preprocessing, and inference-runtime changes. 

Ways to stand out from the crowd: 

• Experience delivering end-to-end video AI applications, including video ingestion, inference, tracking, analytics, and downstream integrations, using technologies such as RTSP and FFmpeg. 

• Background with NVIDIA DeepStream, TensorRT, Triton Inference Server, CUDA, or comparable technologies. 

• Experience optimizing and profiling GPU inference pipelines, including batching, memory management, and concurrent model execution. 

• Experience with stateful streaming inference, KV-cache management, batching, quantization, and their impact on serving performance. 

• Experience profiling GPU applications with tools such as NVIDIA Nsight Systems and deploying inference services across x86 and ARM platforms.  

NVIDIA is widely considered to be one of the technology world’s most desirable employers. We have some of the most forward-thinking and passionate people in the world working for us. Are you a creative problem solver with a passion for solving real-world problems with AI? If so, we want to hear from you.