Machine Learning Engineer

At Sign-Speak Inc, the Machine Learning team builds and deploys AI systems that support sign language recognition and communication. This onsite role in Rochester, NY is focused on developing high-performing models, accelerating training and inference with GPUs, and bringing production-ready solutions to both real-time applications and local devices.

What you will do

• Develop, train, evaluate, and iterate on machine learning models powering sign language recognition and communication systems.
• Implement and experiment with machine learning architectures and techniques from research papers.
• Optimize models for accuracy, latency, throughput, and computational efficiency.
• Build and improve GPU-accelerated training and inference pipelines.
• Apply performance techniques such as model quantization, pruning, and other optimization methods.
• Deploy and maintain models in production environments.
• Develop and optimize models for execution on local devices, including smartphones and other edge computing platforms.
• Integrate machine learning models into real-time applications.
• Improve machine learning infrastructure, model serving, and deployment workflows.
• Investigate performance bottlenecks related to execution, memory utilization, and hardware acceleration.
• Evaluate accuracy versus inference speed tradeoffs and compute resource requirements.
• Write and execute technical design documents.
• Collaborate with research and engineering teams to transition experimental work into production-ready solutions.

Qualifications

• Degree in Computer Science, Machine Learning, Artificial Intelligence, Computer Engineering, or equivalent professional experience.
• Preferred PhD.
• Experience developing, training, and evaluating machine learning models.
• Proficiency in Python and familiarity with common machine learning frameworks.
• Experience implementing or adapting machine learning techniques from research papers.
• Understanding of neural networks, deep learning architectures, and model optimization.
• Experience deploying machine learning models in production or other performance-constrained environments.
• Strong understanding of GPUs and hardware-accelerated machine learning.
• Familiarity with model inference optimization and performance benchmarking.
• Understanding of software engineering principles, backend systems, and machine learning pipelines.
• Willingness to learn and work in a fast-paced startup environment.

Tools and technologies

• Python, PyTorch, TensorFlow
• CUDA
• ONNX Runtime, TensorRT
• Core ML, TensorFlow Lite, LiteRT
• MLOps, model serving, and deployment pipelines

Compensation and benefits

Base salary: $120,000 to $160,000 annually.

• Health insurance
• Dental insurance
• Vision insurance
• Paid time off
• Professional development support
• Stock options

Preferred areas of technical experience

• GPU computing and optimization, including CUDA
• Deep learning training, fine-tuning, and evaluation
• Model quantization, pruning, distillation, and compression
• Inference optimization using ONNX Runtime or TensorRT
• On-device deployment with Core ML, TensorFlow Lite/LiteRT, or similar frameworks
• Real-time, low-latency AI applications
• Computer vision, video understanding, and human pose estimation
• Sequence modeling and transformer-based architectures
• Distributed training and cloud-based GPU infrastructure
• Performance profiling and memory optimization
• Knowledge of the Deaf and Hard of Hearing (D/HH) community is always preferred; if you are unfamiliar, Sign-Speak will help you learn

Location, equipment, and work setup

• Required: Living in Upstate New York (examples include Syracuse, Buffalo, or Rochester).
• Preferred: In-person in Rochester, New York.
• This role requires regular use of a computer and related equipment, plus a reliable internet connection.
• Sign-Speak will provide a work-issued laptop, monitor, and standard workstation equipment.
• Additional development equipment or access to specialized computing resources will be provided based on the role’s needs.

Work authorization and additional notes

• Sign-Speak sponsors H1B transfer, but does not support new H1B sponsorship at this time.
• People with a criminal record are encouraged to apply.
• All requirements are subject to possible modification to reasonably accommodate individuals with disabilities.
• This description does not create an employment contract; employment remains at-will.