NVIDIA | vLLM + SGLang | Deep Learning Inference | Remote (North America preferred)
Hi everyone — I’m Akbar, Senior Manager of Deep Learning Inference Software at NVIDIA. I lead our engineering efforts around vLLM and SGLang, two of the most widely used open-source LLM inference frameworks.
We’re building teams focused on making LLM inference faster, more efficient, and more reliable at scale — from runtime and scheduling optimizations to kernel fusion, distributed serving, and continuous integration across new GPU architectures (Hopper, Blackwell, etc.).
These roles are remote-friendly (North America preferred) and fully focused on upstream open-source development — working directly with the maintainers and the wider AI community.
If you’re excited about large-scale inference, compiler/runtime performance, and pushing GPUs to their limits, we’d love to talk.
We believe AI will fundamentally transform how people live and work. CentML's mission is to massively reduce the cost of developing and deploying ML models so we can enable anyone to harness the power of AI and everyone to benefit from its potential.
CentML ([https://centml.ai](https://centml.ai/)) | Director of Engineering (ML Inference), Developer Evangelist (ML Tools) | Full-time | US (San Francisco / Bay Area), Canada (Toronto)
CentML is on a mission on making modern ML affordable for everyone. Our origin and expertise is in machine learning systems space meaning we know a lot about computer architecture, compilers and ML frameworks. Our big insight is that there is a significant opportunity for system level performance tuning of the workloads so that we can alleviate the current GPU shortage by allowing to utilize mid-tier GPUs (i.e. A10 instead of A100s for inference) without sacrificing on performance.
I am one of the co-founders and I am focused on hiring the following two critical technical roles:
- Director of Engineering to lead our Inference Service. We’re looking for someone with an experience deploying large scale ML workloads and experience with compiler technologies.
- Developer Evangelist - for our machine learning tools, DeepView - https://docs.centml.ai/. DeepView is a profiler that we build specifically for ML practitioners to help them visualize the performance bottlenecks of their models and address them before deploying to production. This often results in significant cost savings. The tools are free and we need someone to help get the message out to the community.
I love the tool. Would you be able to add background blending feature? Basically the use case I thinking of replacing the background by merging the two photos, where main photo is subject and the secondary photo is background. Doing blending in Photoshop is such a pain.
CentML (https://centml.ai) | ML compilers engineer, software systems engineer, front-end engineers | Full-time | Toronto, Canada
CentML is an ML Systems startup with a vision of making ML affordable for everyone, hence the Cent. We leverage our expertise in ML, compilers and computer architecture to achieve significant improvement in GPU utilization which reduces the training time and cost. If this sounds interesting and you are curious how we do it, look up our CEO, Gennady Pekhimenko. Alternatively, drop me a note at my first name (at) centml.ai, I am a co-founder and would love to connect.
CentML is an ML Systems startup with a dream of making ML training affordable for everyone, hence the Cent. We leverage our expertise in ML, compilers and hardware to achieve significant improvement in training time and cost. If this sounds interesting and you want to learn more details, look up our CEO, Gennady Pekhimenko, and his publication history. Alternatively, drop me a note at my first name (at) centml.ai, I am a co-founder and would love to connect.
CentML is an ML Systems startup with a dream of making ML training affordable for everyone, hence the Cent. We leverage our expertise in ML, compilers and hardware to achieve significant improvement in training time and cost. If this sounds interesting and you want to learn more details, look up our CEO, Gennady Pekhimenko, and his publication history. Alternatively, drop me a note at my first name (at) centml.ai, I am a co-founder and would love to connect.
We recently raised a seed round and have over two years runway. We are operating in stealth mode, but our mission is to make ML training affordable for everyone. Our proprietary technology built on top of our unique expertise at the intersection of ML, compilers, and hardware allows our customers to realize 50-90% savings on training their production models.
We're looking to bring onboard a few founding engineers. I am a co-founder if you want to learn more drop me a note at my first name (at) centml.ai
Hi everyone — I’m Akbar, Senior Manager of Deep Learning Inference Software at NVIDIA. I lead our engineering efforts around vLLM and SGLang, two of the most widely used open-source LLM inference frameworks.
We’re building teams focused on making LLM inference faster, more efficient, and more reliable at scale — from runtime and scheduling optimizations to kernel fusion, distributed serving, and continuous integration across new GPU architectures (Hopper, Blackwell, etc.).
We’re hiring for multiple roles:
• Senior Deep Learning Software Engineer, Inference (https://nvidia.wd5.myworkdayjobs.com/NVIDIAExternalCareerSit...)
• Engineering Manager, Deep Learning Inference (https://nvidia.wd5.myworkdayjobs.com/NVIDIAExternalCareerSit...)
• DL Performance Software Engineer - LLM Inference (https://nvidia.wd5.myworkdayjobs.com/en-US/NVIDIAExternalCar...)
• DL Performance Software Engineer - LLM Inference (https://nvidia.wd5.myworkdayjobs.com/en-US/NVIDIAExternalCar...)
These roles are remote-friendly (North America preferred) and fully focused on upstream open-source development — working directly with the maintainers and the wider AI community.
If you’re excited about large-scale inference, compiler/runtime performance, and pushing GPUs to their limits, we’d love to talk.