About the job
Skydio stands at the forefront of the drone industry, pioneering autonomous flight technology that is shaping the future of aerial mobility. Our team excels in the realms of artificial intelligence, cutting-edge hardware and software development, and a relentless focus on customer satisfaction. We aim to broaden the accessibility of drone technology to a diverse range of users, from utility inspectors to first responders and military personnel.
Our mission is to harness transformative AI to create unparalleled flying machines for both enterprise and governmental use. Central to our autonomy system is the ability to understand the world through semantic and geometric insights derived from visual data. We are continually pushing the limits of real-time deep learning networks to enhance the capabilities of intelligent mobile robots.
About the Role:
If you’re passionate about utilizing extensive structured video data to tackle challenges in Computer Vision (CV)—including object detection, tracking, optical flow estimation, and segmentation—we would be thrilled to connect with you.
How You’ll Make an Impact:
Develop high-performance deep learning inference solutions for CV tasks, ensuring high throughput and minimal latency across various hardware platforms.
Profile and analyze CV and Vision Language Models (VLMs) to identify performance bottlenecks, optimization opportunities, and enhance power efficiency in deep learning inference workloads.
Design and implement comprehensive MLOps workflows for model deployment, monitoring, and re-training.
Apply advanced Machine Learning expertise to leverage training/runtime frameworks or model efficiency tools to elevate system performance.
Create innovative approaches to enhance training efficiency.
Implement GPU kernels for custom architectures and optimized inference.
Design and develop SDKs that empower customers and external developers to create autonomous workflows using machine learning.
Leverage your expertise to uphold and enhance Skydio’s engineering standards.

