Pay
$90K–$150K is typical
The middle half of what similar US postings state. This one doesn’t list pay.
From 23,355 postings in the last year. Stated ranges, not salaries paid.
San Francisco, California, United States
Open 5 months, first seen by us 166 days agoWork setup in the employer’s posting: On-siteFrom the employer’s posting
Employment type in the employer’s posting: FullTimeFrom the employer’s posting
Find and fix bottlenecks in the video analytics pipeline on server and edge hardware
Apply quantization, pruning and layer fusion to ViT and CNN models
Write custom optimized ML operator libraries across the framework and compiler stack
Reduce compute cost per video stream to scale the SaaS product
Build and run systems for collecting, labeling and using visual data for training
Identifying key bottlenecks in our current video analytics pipeline
Applying advanced model optimization techniques—such as quantization (Int8/FP16), pruning, and layer fusion—to our Vision Transformers (ViTs) and CNNs
Working across the entire ML framework/compiler stack (e.g., PyTorch, CUDA, TensorRT, and NVIDIA DeepStream) to write custom optimized ML operator libraries
Reducing the compute cost per video stream to enable massive scalability of our SaaS product
Building, improving, maintaining, and operating systems to facilitate the collection, labeling, and use of visual data for ML training
Experience with model optimization, quantization, and efficient deep learning techniques (e.g., knowledge distillation, pruning).
Deep understanding of GPU hardware performance, including execution models, thread hierarchy, memory/cache management, and the cost/performance trade-offs of video processing.
Experience with profiling and benchmarking tools (e.g., Nsight Systems, Nsight Compute) to validate performance on complex architectures.
Experience identifying and resolving compute and data flow bottlenecks, particularly in high-bandwidth video processing pipelines.
Strong communication skills and the ability to work cross-functionally between research and infrastructure teams.
Familiarity with database systems (e.g., SQL, Neo4j).
Experience with video processing frameworks such as NVIDIA DeepStream, DALI, or FFmpeg.
Familiarity with ML compilers (e.g., TVM, MLIR) or inference engines like TensorRT or ONNX Runtime.
Knowledge of distributed training systems or cloud-scale inference serving (e.g., Triton Inference Server).
Requirements - BS/MS or Ph.D. in Computer Science, Electrical Engineering, or a related discipline. - Strong programming skills in C/C++ and Python.
Open 5 monthsWe first saw this posting 166 days ago and it is still up. Roles open this long are often slow to fill or kept open to collect applications.
These dates show when we found the posting. Check the employer’s website to confirm it is still accepting applications.
$90K–$150K is typical
The middle half of what similar US postings state. This one doesn’t list pay.
From 23,355 postings in the last year. Stated ranges, not salaries paid.
Day 166
Up longer than most. Similar jobs usually come down between day 9 and day 60.
From 40,907 similar postings that came down. Coming down is not always a hire.
2 open jobs
Add your resume and we’ll mark which of these tools you already use, and point out anything in the posting that could rule you out.
- BS/MS or Ph.D. in Computer Science, Electrical Engineering, or a related discipline. - Strong programming skills in C/C++ and Python. - Experience with model optimization, quantization, and efficient deep learning techniques (e.g., knowledge distillation, pruning).
Preferred Qualifications - Familiarity with database systems (e.g., SQL, Neo4j). - Work in Computer Vision, Deep Learning, and Vision Transformers.
- Model Acceleration: Applying advanced model optimization techniques—such as quantization (Int8/FP16), pruning, and layer fusion—to our Vision Transformers (ViTs) and CNNs to maximize throughput and minimize latency. - Building Efficient Operators: Working across the entire ML framework/compiler stack (e.g., PyTorch, CUDA, TensorRT, and NVIDIA DeepStream) to write custom optimized ML operator libraries. - Resource Efficiency: Reducing the compute cost per video stream to enable massive scalability of our SaaS product.
Complete your application on jobs.ashbyhq.com. The employer’s form will show what is required.
Already applied? Track this application
The AI Infrastructure team at Zensors builds the engine that powers our visual sensing platform. We provide the tools to automate the lifecycle of our AI workflow, including model development, evaluation, optimization, deployment, and monitoring across thousands of video streams.
As a Machine Learning Engineer in ML Runtime & Optimization, you will develop technologies to accelerate the training and inference of computer vision models that power smart spaces and cities.
Your responsibilities will include:
Optimizing Core ML Pipelines: Identifying key bottlenecks in our current video analytics pipeline and performing in-depth analysis to ensure the best possible performance on current server and edge compute architectures.
Cross-Stack Collaboration: Collaborating closely with AI research and platform engineering teams to optimize core parallel algorithms and influence the design of our next-generation inference infrastructure.
Model Acceleration: Applying advanced model optimization techniques—such as quantization (Int8/FP16), pruning, and layer fusion—to our Vision Transformers (ViTs) and CNNs to maximize throughput and minimize latency.
Building Efficient Operators: Working across the entire ML framework/compiler stack (e.g., PyTorch, CUDA, TensorRT, and NVIDIA DeepStream) to write custom optimized ML operator libraries.
Resource Efficiency: Reducing the compute cost per video stream to enable massive scalability of our SaaS product.
Data Management: Building, improving, maintaining, and operating systems to facilitate the collection, labeling, and use of visual data for ML training.
BS/MS or Ph.D. in Computer Science, Electrical Engineering, or a related discipline.
Strong programming skills in C/C++ and Python.
Experience with model optimization, quantization, and efficient deep learning techniques (e.g., knowledge distillation, pruning).
Deep understanding of GPU hardware performance, including execution models, thread hierarchy, memory/cache management, and the cost/performance trade-offs of video processing.
Experience with profiling and benchmarking tools (e.g., Nsight Systems, Nsight Compute) to validate performance on complex architectures.
Experience identifying and resolving compute and data flow bottlenecks, particularly in high-bandwidth video processing pipelines.
Strong communication skills and the ability to work cross-functionally between research and infrastructure teams.
Familiarity with database systems (e.g., SQL, Neo4j).
Work in Computer Vision, Deep Learning, and Vision Transformers.
Experience with video processing frameworks such as NVIDIA DeepStream, DALI, or FFmpeg.
Familiarity with ML compilers (e.g., TVM, MLIR) or inference engines like TensorRT or ONNX Runtime.
Knowledge of distributed training systems or cloud-scale inference serving (e.g., Triton Inference Server).
Selected passages from the saved posting. Check the full description for conditions and exceptions.
No pay amount identified in the saved description.
San Francisco, California, United States
Working pattern and location restrictions need checking in the full posting.
No clear work-authorization passage found. Eligibility is unconfirmed.