about 4 hours ago
Bellevue, WA, USAMid Level
H1B Sponsor
Base Salary
$155k - $193k/yr
Responsibilities
- Translate operational problems into AI requirements, datasets, evaluation criteria, and production architectures.
- Design, train, fine-tune, evaluate, and deploy models for vision, language, multimodal AI, time-series analysis, autonomy, and optimization.
- Apply research methods including fine-tuning, distillation, quantization, pruning, batching, caching, and hardware-aware acceleration.
- Build datasets from video, images, text, telemetry, sensor, synthetic, and other multimodal data.
- Evaluate models for accuracy, latency, throughput, robustness, safety, groundedness, and resource efficiency.
- Build reliable Python AI services and containerized workloads across Kubernetes, cloud, on-premises, and disconnected edge environments.
- Develop monitoring, data validation, drift detection, retraining, and controlled-update pipelines.
- Diagnose issues across data pipelines, models, GPUs, orchestration, networking, and applications.
- Partner with customers, product teams, engineers, and domain experts to move prototypes into production.
- Contribute reusable models, datasets, evaluation tools, platform components, documentation, patents, and publications.
Requirements
- Master’s or PhD in applied mathematics, computer science, computational science, engineering, or a related technical field.
- At least three years of industry experience in AI research, machine learning, and production software development.
- Strong Python programming skills and proficiency in Java, C++, or another production language.
- Hands-on experience with statistical machine learning, deep learning, natural language processing, and modern neural architectures.
- Proficiency with PyTorch, TensorFlow, or JAX.
- Familiarity with containers, numerical libraries, modular software design, version control, and automated testing.
- Experience applying machine learning to real-world problems and deploying models beyond research prototypes.
- Evidence of technical depth through peer-reviewed publications, open-source contributions, or significant production systems.
- Preferred experience with GPU and edge deployment, TensorRT, Triton Inference Server, ONNX Runtime, vLLM, CUDA, Kubernetes, distributed systems, CI/CD, production MLOps, multimodal data, robotics, industrial or safety-critical systems, quantization, pruning, distillation, LoRA, synthetic data, active learning, weak supervision, human-in-the-loop workflows, or simulation.
Benefits
- Medical, dental, and vision coverage at subsidized cost.
- Health savings accounts, flexible spending accounts, and dependent care FSAs.
- 401(k) and Roth 401(k) retirement plan options.
- Unlimited paid time off and 14 paid company holidays per year.
- Office-based work at the Bellevue, Washington office.
- Equity is offered in addition to base salary.
About Armada
Welcome to the new edge. Armada is the world’s first full-stack edge computing platform, revolutionizing connectivity, compute, and AI solutions where they’re needed most - anywhere on Earth.