ML Frameworks Engineer

Hailo is a leading high-tech company, developing top-performing AI processors for edge devices in various industries. We combine an extensive understanding of the way neural networks operate, with our team’s expertise in SW and HW architecture, to develop products that transform the way we use machines to perceive and analyze the world around us.

Our company is led by a group of experts from a wide range of disciplines, combining experience from elite intelligence units as well as leading tech companies in Israel and abroad, with a track record of executing complex projects from the ground up. As an AI chip company, we believe that AI can help us create a better, safer, more productive, and more convenient world. For this to happen, AI needs to be available at scale and on the edge.

To this end, our team of hardware and software leaders is developing top-performing AI processors, dedicated to AI tasks on edge devices in various applications and industries including smart cities, automotive, manufacturing, agriculture, retail, and many more. With hundreds of customer programs around the world, and a wide ecosystem of software and hardware partners, Hailo is leading the edge AI revolution.

About the position

As part of the Machine Learning Frameworks team, you’ll be tasked with the development and implementation of a crucial part in Hailo’s SDK (dataflow compiler, aka DFC).

Who Are We Looking For?

We are looking for applicants with an R&D background who are versatile and can learn new skills quickly. The position is system oriented, and we are looking for candidates who will feel comfortable when required to deep dive into our HW architecture, SW toolchain or the customers application. Deep learning development is SW intensive, and applicants are expected to have knowledge and experience of scientific computing and programming. In general, we hold SW and automation skills in high regard. You’re an ideal candidate if you’re curious, independent, and accountable, who enjoys teamwork and knowledge sharing with your peers.

The team oversees:

• Development of Hailo’s model translator which converts deep learning models into Hailo’s internal representation. Common task: detect and replace graph blocks of neural computation operators with mathematical equivalents that can be represented as fused layers.
• Design and development of various high-level APIs, for improved usability of complex capabilities of Hailo’s DFC product. Common task: Add a new CLI command for adding an input format conversion while running an evaluation on a model from Hailo’s model zoo.
• Implement state-of-the-art neural network architectures and operators with awareness of Hailo’s hardware device capabilities. Common task: implement emulation software to support new structures such as vision transformers (e.g., ViT, DETR, etc.) and various post-processes (e.g., non-max-suppression, Yolo, etc.).

• B.Sc. in CS/EE or military equivalent experience.
• Extensive experience development in Python/CPP (minimum of 2 years).
• Hard-working, committed, and self-reliant.

• Experience in Deep learning frameworks like Pytorch and TensorFlow.
• Experience with MLops.
• SW architecture skills.

We are passionate about building an inclusive and equitable working environment.

We promote a flexible work environment that encourages work-life balance.

If you don’t meet 100% of the requirements– no worries!

Our DNA is characterized by openness to multi-disciplinary people, who bring great potential to grow with us, not just a skills setlist. We encourage you to apply if you think this role would excite you about coming to work every day.