Director Embedded Artificial Intelligence

Exosens; Grenoble; Hybrid; EUR 140,000 - 210,000; Full time

Job summary

Exosens in Grenoble, France, seeks a Director of Embedded Artificial Intelligence to build and lead a new team focused on integrating AI with sensing hardware for imaging systems.

You will drive AI strategy across hardware-aware optimization, porting to NN accelerators, FPGAs and GPUs, and deliver production-ready solutions for defense, life sciences and industrial sensing.

This role combines hands-on technical leadership with strategic roadmap, performance, power, and cost targets.

Qualifications

• Proven experience deploying AI or advanced signal processing on embedded hardware in commercial products.
• Strong practical expertise in porting and optimizing algorithms for dedicated neural network accelerators, FPGAs, or GPUs.
• Solid understanding of embedded computing, parallel processing, fixed-point arithmetic, memory hierarchies, and hardware constraints.
• Experience with hardware-specific compilation, inference runtimes, profiling tools, and debugging on target devices.
• Strong C/C++ and Python skills with modern deep learning frameworks.

Responsibilities

• Build and lead a team combining expertise in AI algorithms, embedded software, and hardware acceleration.
• Define the AI architecture and technology roadmap for camera, sensor, and detector products.
• Lead porting, optimization, and deployment of neural networks and signal processing on NN processors, FPGAs, GPUs, and embedded SoCs.
• Guide hardware-software co-design including processor selection, workload partitioning, memory architecture, and data movement.
• Adapt models to target hardware through quantization, pruning, and custom operator implementations.
• Develop efficient inference pipelines from sensor acquisition to postprocessing.

Skills

Embedded AI

AI algorithms porting

C/C++

Python

Hardware acceleration

Team leadership

Computer vision

Sensor signal processing

Education

Master's degree or PhD in EE/CS or related

Tools

FPGA toolchains

CUDA

RTL

High-level Synthesis

NN processors

GPUs

Job description

Director Embedded Artificial Intelligence

We are seeking a Director embedded Artificial Intelligence to build and lead a new team dedicated to bringing and integrating (Physical) AI into our next generation of imaging systems, detectors, sensors, cameras and imaging systems. Our focus is on developing intelligent devices that perceive and interpret the physical world through AI tightly integrated with sensing hardware and embedded computing for advanced sensing in Defence, life sciences, industrial control.

This is a hands-on technical leadership role with a strong emphasis on embedded AI and the deployment of algorithms onto silicon , including dedicated neural network processors, FPGAs, and GPUs. You will lead the transition from algorithm development to efficient, reliable execution on production hardware, enhancing the sensing and imaging capabilities of sensors in relation to advanced applications (like anti-drone detection, integrated soldier, multi-spectral analysis, Semiconductor control sensing, non-destructive testing, etc.)

Your mission

Propose, Define and deliver our embedded AI strategy, bringing together algorithm development, hardware-aware optimization, and system integration. You will ensure that AI capabilities meet demanding requirements for detection performance, real-time processing, power consumption, memory, and product cost.

Key responsibilities

• Build and lead a team combining expertise in AI algorithms, embedded software, and hardware acceleration.
• Define the AI architecture and technology roadmap for camera, sensor, and detector products.
• Lead the porting, optimization, and deployment of neural networks and signal processing algorithms onto dedicated NN processors, FPGAs, GPUs, and embedded SoCs.
• Guide hardware-software co-design, including processor selection, workload partitioning, memory architecture, and data movement.
• Adapt models to target hardware through quantization, pruning, compression, hardware-supported operator selection, and custom operator implementation where needed.
• Develop efficient inference pipelines, including sensor acquisition, pre-processing, inference, and postprocessing.
• Profile and resolve bottlenecks in latency, throughput, memory bandwidth, power consumption, and thermal performance.
• Work closely with electronics, FPGA, firmware, and software teams to integrate AI into complete products.
• Establish validation methods covering numerical accuracy after porting, detection performance, timing, and robustness under real operating conditions.
• Own the development lifecycle from feasibility studies and prototypes through production deployment and maintenance.
• Evaluate silicon vendors, toolchains, inference runtimes, and external technology partners.

Essential qualifications and experience

• Proven experience deploying AI or advanced signal processing algorithms on embedded hardware in commercial products.
• Strong practical expertise in porting and optimizing algorithms for dedicated neural network accelerators, FPGAs, or GPUs , with deep expertise in at least one and the ability to lead development across these architectures.
• Solid understanding of embedded computing, parallel processing, fixed-point arithmetic, memory hierarchies, and hardware resource constraints.
• Experience with hardware-specific compilation, inference runtimes, profiling tools, and debugging on target devices.
• Strong C/C++ and Python skills, with experience using modern deep learning frameworks.
• Expertise in computer vision, image processing, or sensor signal processing.
• Demonstrated ability to balance model accuracy with real-time performance, power, memory, and cost.
• Proven technical leadership and experience building or managing multidisciplinary engineering teams.
• Master's degree or PhD in electrical engineering, computer engineering, computer science, or a related field, or equivalent practical experience.

Additional experience valued

• FPGA implementation using RTL and/or high-level synthesis.
• GPU kernel development and optimization.
• Quantization-aware training, mixed-precision inference, and numerical validation.
• Embedded Linux, RTOS environments, device drivers, and sensor interfaces.
• Multi-sensor processing, sensor fusion, and non-visible-spectrum imaging.
• Taking embedded AI products through qualification and volume production.

Position based in Grenoble, France.