Expert DL Deployment Engineer IRC283851

Description

Our client is dedicated to making safe, intelligent mobility a reality. Headquartered in Sweden, they develop a complete, scalable software stack for ADAS and autonomous driving—from sensing to actuation—and are soon to launch with a major global automaker.

Driven by the goal of reducing fatal accidents by 100%, their team of 500+ global engineers combines technical excellence with a shared passion for saving lives. The company values a T-shaped competence profile, prioritizing a strong mindset, responsibility, and collaboration over a pure skill set. They offer a flexible hybrid work environment where ambition, mentorship, and software craftsmanship thrive.

Requirements

To be considered an “Expert” for this role, you must demonstrate a proven track record of shipping neural networks to target hardware in previous professional environments.

Core Technical Competencies

High-Performance Computing (C++ & Python):

• Expert-level proficiency in C++ for low-latency production environments.
• Strong command of Python for model interfacing, scripting, and automation.
• Solid software engineering background (OOP, design patterns, version control) with a focus on writing clean, maintainable, and testable code.

GPU & Parallel Programming:

• Deep understanding of GPU architectures and memory management.
• Hands-on experience with CUDA programming, including writing and optimizing custom kernels to squeeze maximum performance out of the hardware.

Inference Optimization:

• Extensive experience with NVIDIA TensorRT for optimizing deep learning models.
• Proficiency in model compression techniques, including quantization (INT8/FP16), pruning, and layer fusion.
• Familiarity with model exchange formats (e.g., ONNX) and how to debug conversion failures.

Production & Embedded Engineering

• Deployment Strategy: Experience deploying software in embedded environments or edge devices (e.g., NVIDIA Jetson, Orin, or custom hardware).
• Resource Constraints: Ability to work within strict constraints regarding power consumption, thermal limits, and memory bandwidth without sacrificing model accuracy.
• Production Readiness: Experience integrating inference engines into larger production pipelines, ensuring robustness, stability, and error handling.

The “Expert” Factor (Soft Skills & Leadership)

• Architectural Ownership: Ability to make high-level decisions regarding hardware selection and software stack architecture.
• Mentorship: Willingness to guide junior engineers and share knowledge on optimization best practices.
• Problem Solving: Demonstrated ability to debug complex issues at the intersection of hardware, drivers, and software.

Ideally, you also have… (Bonus Qualifications)

• Knowledge of Docker/Containerization for reproducible deployment.
• Familiarity with CI/CD pipelines specifically for Machine Learning (MLOps).
• Background in Computer Vision or Signal Processing.

#LI-LY1 #LI-Onsite 

Job responsibilities

Job Responsibilities
As a key team member, you will bridge the gap between experimental deep learning research and our production codebase. You will own the lifecycle of model deployment, ensuring our neural networks run at peak efficiency on our target hardware.

Deployment & Compilation

• Model Transition: Lead the deployment of deep learning models from the research environment into the production product code.
• Engine Generation: Compile and generate optimized TensorRT engine files tailored for specific target platforms, including x86, NVIDIA Xavier, and Orin.
• Version Management: Manage the versioning and compatibility of deployed models within the larger system architecture.

Custom Kernel Development

• Gap Analysis: Identify limitations in NVIDIA DriveOS layer support required for next-gen DL models.
• Plugin Implementation: Design, code, and integrate custom CUDA kernels as TensorRT plugins to enable unsupported network layers.
• High-Performance Coding: Write highly optimized C++ code to bridge the gap between research model architectures and hardware capabilities.

Performance Optimization

• Efficiency Tuning: Drive the optimization strategy to ensure models run efficiently on System-on-Chip (SoC) constraints.
• Advanced Techniques: Apply and refine hardware-aware optimization techniques.

What we offer

Culture of caring. At GlobalLogic, we prioritize a culture of caring. Across every region and department, at every level, we consistently put people first. From day one, you’ll experience an inclusive culture of acceptance and belonging, where you’ll have the chance to build meaningful connections with collaborative teammates, supportive managers, and compassionate leaders. 

Learning and development. We are committed to your continuous learning and development. You’ll learn and grow daily in an environment with many opportunities to try new things, sharpen your skills, and advance your career at GlobalLogic. With our Career Navigator tool as just one example, GlobalLogic offers a rich array of programs, training curricula, and hands-on opportunities to grow personally and professionally.

Interesting & meaningful work. GlobalLogic is known for engineering impact for and with clients around the world. As part of our team, you’ll have the chance to work on projects that matter. Each is a unique opportunity to engage your curiosity and creative problem-solving skills as you help clients reimagine what’s possible and bring new solutions to market. In the process, you’ll have the privilege of working on some of the most cutting-edge and impactful solutions shaping the world today.

Balance and flexibility. We believe in the importance of balance and flexibility. With many functional career areas, roles, and work arrangements, you can explore ways of achieving the perfect balance between your work and life. Your life extends beyond the office, and we always do our best to help you integrate and balance the best of work and life, having fun along the way!

High-trust organization. We are a high-trust organization where integrity is key. By joining GlobalLogic, you’re placing your trust in a safe, reliable, and ethical global company. Integrity and trust are a cornerstone of our value proposition to our employees and clients. You will find truthfulness, candor, and integrity in everything we do.

About GlobalLogic

GlobalLogic, a Hitachi Group Company, is a trusted digital engineering partner to the world’s largest and most forward-thinking companies. Since 2000, we’ve been at the forefront of the digital revolution – helping create some of the most innovative and widely used digital products and experiences. Today we continue to collaborate with clients in transforming businesses and redefining industries through intelligent products, platforms, and services.