Edge AI Engineer (f/m/d)

At NXAI, we're always excited to meet forward-thinking minds eager to shape the future of AI. We work at the cutting edge of technology with a team of technical experts to continue creating excellent solutions.

In this role, you will ensure ML models run efficiently by developing, optimizing, and adapting AI models for edge devices. Your primary duties will focus on:

• Hardware-Specific Deployment: Deploy and scale AI models across a diverse hardware fleet, including NVIDIA Jetson, FPGAs, TPUs, Raspberry Pi, and Industrial PCs.

• Performance Engineering: Convert and optimize models for edge runtimes to maximize throughput, utilizing quantization, pruning, and mixed-precision to fit models into resource-constrained environments.

• Low-Latency Architecture: Develop robust system architectures that ensure real-time data processing and seamless integration of AI models into edge devices.

• Kernel & Inference Optimization: Develop and optimize CUDA kernels (specifically for xLSTM architectures) and manage hardware resources to ensure ultra-low latency inference.

• Performance Profiling: Deep-dive into the stack to profile and debug performance bottlenecks, ensuring peak efficiency of hardware-accelerated pipelines.

• Architectural Collaboration: Bridge the gap between research and production, collaborating with the research team to build scalable, fast, and secure AI architectures that prioritize data privacy and system integrity at the edge.

• End-to-End Integration: Manage complex data flows and integrate security mechanisms to protect models and privacy on decentralized devices.

You may be a good fit if you demonstrate excellence across these core competencies:

• Software Mastery: Strong proficiency in Python and expert-level knowledge of C/C++ for building low-latency inference pipelines. Familiarity with Rust or C for safety-critical systems is a significant advantage.

• Optimization Expert: An in-depth understanding of quantization, pruning, mixed-precision, and TensorRT optimizations to make models both memory-efficient and lightning-fast.

• Frameworks & Runtimes: Solid hands-on experience with PyTorch and ONNX Runtime for deploying models across diverse edge device architectures and NPU/TPU accelerators (e.g., Google Coral).

• Hardware Acceleration: A proven track record of enabling models to run on embedded platforms such as NVIDIA Jetson, ARM-based systems (SIMD/NEON), or microcontrollers, including experience with GPU inference and hardware acceleration (e.g., AMD ROCm).

• Embedded Systems: Proficient knowledge of (Embedded) Linux systems, including driver debugging and working with distributions like Yocto or Ubuntu for ARM.

• Mindset: Exceptional problem-solving skills and the ability to work independently in a fast-paced environment, taking ownership of the full deployment stack.

What you'll get:

• We offer you the opportunity to work in a key position at the forefront of a fast-growing company.

• Competitive salary in line with the market, plus a tailored benefits package.

• Benefit from a diverse and probably one of the best teams in the field of new evolving AI technology, where professional excellence is the guiding principle.

• The chance to make a real impact and shaping the future of AI.

• A market-aligned salary package tailored to your experience, plus individual benefits. Starting salary is €3,900 gross per month. Depending on your profile and your development potential within the team, a significant overpayment is of course possible. The salary is paid 14 times per year.