Principal AI Engineer

Valeo; Tuam; On-site; EUR 120,000 - 180,000; Full time

Job summary

Valeo anSWer Product Line in Ireland is seeking a Principal AI Engineer to lead the architectural design of end-to-end deep learning solutions for automated driving and parking. You will guide cross-disciplinary teams to develop holistic sensor-to-action AI models and shape state-of-the-art E2E algorithms using large driving datasets.

You will push advances in LDMs, embodied AI, and edge compute optimization across multi-core hardware, while aligning with safety standards and pursuing IP

Qualifications

• 7+ years of applied deep learning and autonomous systems experience.
• Expertise in Vision Transformers, BEV networks, and Vision-Language-Action models.
• Experience with edge AI hardware and deploying models to embedded platforms.

Responsibilities

• Architect E2E Driving Models: lead research and design for differentiable sensor-to-control networks.
• Direct AI Governance: chair the AI Steering Committee to align architectures.
• Guide HW-SW Co-Design: inform SoC partitioning and AI optimization for multi-core hardware.
• Deliver technical architectures for pre-development programs and bids.
• Support safety/regulatory alignment and model interpretability considerations.
• Drive research partnerships and IP disclosures for patents.

Skills

Deep Learning

Autonomous Systems

Vision Transformers

Edge AI

NPUs

Embedded Deployment

ML Data Pipelines

Education

BSc/BEng/PhD in Math/Physics/CS/Engineering

Tools

PyTorch

MLOps

Job description

Valeo anSWer Product Line

Valeo anSWer Product Line is a world leader in providing advanced driver-assistance (ADAS) and automated driving/parking systems. The AI Platform team is responsible for developing custom-built, state-of-the-art neural architectures designed to leverage the full power of today's most advanced multi-core automotive embedded platforms and Neural Processing Units (NPUs). Moving beyond traditional modular perception pipelines, the team applies state-of-the-art End-to-End (E2E) deep learning, imitation learning, and reinforcement learning to directly map multi-modal sensor inputs to vehicle control and trajectory planning. We specialize in redefining the state-of-the-art in AI, leveraging large-scale data and foundational driving models to deliver next-generation, fully differentiable automated driving systems for leading global OEMs.

Position Overview

As a Principal AI Engineer, you will be responsible for the architectural design of end-to-end deep learning solutions enabling the next generation of automated driving and parking platforms. Working as part of a multidisciplinary team, you will utilize deep domain knowledge to technically guide the development of holistic, sensor-to-action AI models. You will be responsible for the design of state-of-the-art E2E algorithms, leveraging substantial driving datasets, behavioral cloning, and closed-loop simulation. You will be expected to remain at the absolute forefront of technological developments in the rapidly expanding fields of large driving models, embodied AI, Vision-Language-Action (VLA) models, and edge compute optimization.

Responsibilities

• Architect E2E Driving Models: Lead research and design for fully differentiable, sensor-to-control neural networks powering automated driving and parking.
• Direct AI Governance & Alignment: Chair the AI Technical Steering Committee to approve designs, unify network backbones, and guide legacy algorithm migration.
• Guide Hardware-Software Co-Design: Partner with embedded engineering teams to inform SoC partitioning, pipelining, and AI optimization strategies (quantization, pruning, memory bandwidth) for multi-core ARM, DSP, and dedicated NPU hardware.
• Deliver RFI/RFQ Technical Architectures: Spearhead novel AI system proposals for pre-development programs and commercial vehicle bids.
• Support Safety & Regulatory Compliance: Collaborate with systems and safety engineering to align E2E architectures with automotive safety standards (e.g., ISO 21448 SOTIF) and AI regulations, advising on model interpretability and safety boundaries.
• Drive Research & Intellectual Property: Lead university research partnerships and generate invention disclosure memos to secure strategic patents.

Required Skills/Experience

• Relevant BSc, BEng, or PhD (Mathematics, Physics, Computer Science, or Engineering) with 7+ years of applied experience in deep learning and autonomous systems.
• Deep understanding of modern machine learning architectures (Vision Transformers, Sequence-to-Sequence models, BEV networks, and Vision-Language-Action models).
• Algorithm development experience in end-to-end autonomous driving, imitation learning, reinforcement learning, or advanced trajectory planning.
• Extensive experience with edge AI hardware, heterogeneous SOCs, and Neural Processing Units (NPUs) for deploying large-scale models.
• Proficiency in deploying AI models to embedded platforms.
• Experience with large-scale ML data pipelines, auto-labeling, and driving data curation.
• Proven track record in the automotive, robotics, or autonomous vehicle industry.
• Excellent English language communication skills, both written and verbal.
• High level of innovation, motivation, and technical leadership.
• Academic publications and/or active IP/patents in the fields of deep learning, automated driving, or robotics.

Desired Attributes

• Experience with closed-loop autonomous driving simulators (e.g., CARLA, internally developed simulation engines) and neural simulators (e.g., NeRFs, 3D Gaussian Splatting for driving).
• Deep understanding of modern ML frameworks (PyTorch 2.x) and MLOps pipelines for continuous training of E2E models.