About The Company
Zendar is pioneering the development of a radar-centric autonomy stack designed to enable vehicles, from cars to robots, to operate autonomously across diverse environments. Leveraging deep radar technology, Zendar's solutions place RF sensing at the core of perception systems, allowing for robust detection and navigation capabilities even in challenging conditions such as long-range scenarios, high speeds, and adverse weather. The company's innovative approach results in highly efficient, cost-effective autonomous systems that require minimal hardware investment, making autonomous vehicle deployment more accessible and scalable. With an end-to-end development process that spans radar hardware, signal processing, multi-modal perception models, and path planning, Zendar offers a comprehensive platform that integrates hardware and software seamlessly. The company's mission is to bridge the gap between research innovations and real-world autonomous applications, providing solutions that are both technically advanced and economically viable.
About The Role
We are seeking experienced Machine Learning (ML) Engineers to join our team and play a pivotal role in optimizing and deploying machine learning models on heterogeneous embedded computing platforms. In this role, you will operate at the intersection of machine learning, compilers, runtime systems, and computer architecture, helping to translate research-developed models into highly optimized, production-ready implementations. Your primary focus will be understanding the trade-offs between model quality and computational efficiency, working closely with research teams to develop hardware-aware neural architectures, identify bottlenecks, and explore architectural modifications that enhance latency, throughput, and memory usage without compromising accuracy.
The ideal candidate will have a passion for understanding neural network architectures and the underlying hardware, with expertise in techniques such as hardware-aware neural architecture search, model scaling, quantization, mixed-precision inference, and model compression. This is an exciting opportunity to contribute to real-world challenges by bringing lab-developed algorithms into physical environments, particularly in automotive and robotic applications where performance and reliability are critical.
Qualifications
- Strong understanding of machine learning and deep neural network architectures
- Hands-on experience developing machine learning models using frameworks such as PyTorch
- Proficiency in programming with Python
- Experience analyzing the computational characteristics of neural networks and their inference performance
- Knowledge of techniques including model architecture search, model scaling, quantization, mixed-precision inference, and model compression
- Experience with inference deployment frameworks such as ONNX, TensorRT, or similar
- Ability to reason across different layers of the ML deployment stack, from models to hardware execution
- Familiarity with software development practices including Git, debugging, profiling, and unit testing
- Strong communication skills and ability to collaborate across research, software, and engineering teams
Responsibilities
- Profile and analyze machine learning models to identify bottlenecks in computation, memory, and data movement
- Explore and optimize the trade-offs between model output quality and computational efficiency, focusing on latency, throughput, and memory footprint
- Develop methodologies for hardware-aware model optimization and neural architecture search, utilizing real hardware measurements as optimization objectives
- Apply model optimization techniques such as quantization, mixed-precision inference, and model compression, analyzing their impact on numerical accuracy
- Develop and maintain model export, benchmarking, and deployment pipelines across frameworks like PyTorch, ONNX, and TensorRT
- Evaluate deployment strategies and determine optimal mapping of models onto heterogeneous processing units including CPUs, GPUs, and dedicated AI accelerators
- Collaborate with research teams to iterate on model architectures and optimization techniques for embedded hardware platforms
Benefits
- Opportunity to impact a young, venture-backed company operating in an emerging market
- Competitive salary ranging from €75,000 to €90,000 annually, depending on experience and equity
- Hybrid work model with in-office presence three days a week (Monday, Tuesday, Thursday) and flexible remote work
- Modern, fully equipped workspace located in the heart of Paris
- Transportation benefits including partial reimbursement for public transit
- Subsidized meal vouchers (tickets restaurant)
- Wellness allowance through programs like Gymlib
Equal Opportunity
Zendar is committed to fostering a diverse and inclusive environment. We are proud to be an equal opportunity employer and welcome applications from all qualified candidates regardless of race, color, religion, gender, gender identity or expression, sexual orientation, national origin, genetics, disability, age, or veteran status. All qualified applicants will receive consideration for employment without discrimination.