5 months ago
Zürich, SwitzerlandSenior
Responsibilities
- Develop high-performance, high-throughput, low-latency deep-learning inference solutions for computer-vision workloads across hardware platforms.
- Profile computer-vision and vision-language models to identify bottlenecks, optimization opportunities, and power-efficiency improvements.
- Design and implement end-to-end MLOps workflows for model deployment, monitoring, and retraining.
- Use machine-learning knowledge, training and runtime frameworks, and model-efficiency tools to improve system performance.
- Create methods to improve training efficiency.
- Implement GPU kernels for custom architectures and optimized inference.
- Design and implement SDKs enabling customers and external developers to create autonomous machine-learning workflows.
- Improve engineering standards and collaborate across autonomy, embedded, and cloud teams.
Requirements
- Hands-on experience with MLOps, machine-learning inference acceleration and optimization, and edge deployment.
- Strong knowledge of deep-learning fundamentals, techniques, and current deep-learning model architectures.
- Strong fundamentals in computer vision, image processing, and video processing.
- Hands-on experience building and managing machine-learning pipelines for vision or vision-language tasks, including data preparation, training, deployment, and monitoring.
- Experience with security and compliance requirements in machine-learning infrastructure.
- Experience with machine-learning frameworks and libraries.
- Ability to drive concepts through architecture, development, testing, deployment, and monitoring.
- Ability to navigate and deliver within a complex codebase.
- Strong communication and collaboration skills across technical levels.
Categories
About Skydio
Skydio builds autonomous drones, docking systems, and fleet management software for enterprise, public safety, and defense customers. It sells hardware with AI-based autonomy plus SaaS for fleet operations, inspections, and Drone-as-First-Responder programs. Founded in 2014 and headquartered in San Mateo, California, the company designs and assembles its products in the U.S., with manufacturing in Hayward, and its systems are used across utilities, law enforcement, and the U.S. military.