
Job Overview
Location
Remote - USA
Job Type
Full-time
Category
Software Engineering
Date Posted
July 4, 2026
Full Job Description
đź“‹ Description
- • Design and train large-scale transformer and hybrid foundation models for automotive AI applications, focusing on text, multimodal, and emerging paradigms.
- • Own model architecture decisions across attention variants including RoPE, ALiBi, Grouped Query Attention (GQA), and Mixture-of-Experts (MoE), building models from first principles rather than adapting existing codebases.
- • Diagnose and resolve training instabilities at scale, including optimizer divergence, gradient pathologies, and system-level failures during large-model training.
- • Navigate scaling tradeoffs between data, compute, and model architecture using validated scaling laws and Chinchilla-style optimization principles.
- • Select and implement optimization strategies including AdamW, Lion, and Adafactor optimizers, alongside learning-rate and warmup schedulers to ensure stable convergence.
- • Design and experiment with loss functions such as next-token prediction, contrastive objectives, RLHF, DPO, and GRPO to improve model alignment and generalization.
- • Execute distributed training at scale using FSDP, ZeRO-3, tensor parallelism, pipeline parallelism, mixed precision (bf16, fp8), and gradient checkpointing.
- • Partner with ML systems teams while retaining full architectural ownership of foundation models, ensuring efficient inference with KV cache optimization.
- • Explore and implement novel architectures including MoE routing strategies, multimodal fusion, and SSM/hybrid models with explicit consideration for real-world deployment constraints.
- • Drive technical direction for next-generation AI models in transportation, ensuring models converge faster, generalize better, and exhibit predictable failure modes.
- • Establish in-house expertise in foundation model development, moving beyond adaptation to original research and engineering of scalable AI systems for automotive use cases.
- • Solve critical problems such as why training diverges at scale, how optimizer dynamics interact with architecture, when scaling laws break down, and the real tradeoffs between data, compute, and model design.
- • Maintain rigorous standards for principled, defensible architectural decisions that directly impact model performance, stability, and deployment efficiency in production vehicle systems.
🎯 Requirements
- • Deep theoretical and practical understanding of modern deep learning
- • Hands-on experience training large models from scratch
- • Ability to reason about optimization, not just tune hyperparameters
- • Comfort operating in ambiguous, research-driven environments
- • Strong knowledge of Transformer internals and attention mechanisms
- • Expertise in optimization algorithms, training dynamics, scaling laws, distributed training, and mixed precision
🏖️ Benefits
- • Salary range of $185,000.00 - $280,000.00
- • Annual bonus opportunity
- • Insurance coverage (medical, dental, vision, life, and disability)
- • Paid time off and paid holidays
- • Company contribution to the RRSP (Registered Retirement Savings Plan)
- • Equity awards for certain positions and levels
Skills & Technologies
See exactly how your profile matches this role — strengths, skill gaps, and what to do about them.
About Cerence Inc.
Cerence Inc. develops AI-powered mobility assistant technologies, including voice, natural language understanding, and human-machine interaction software. The company supplies automakers and mobility OEMs with embedded and cloud solutions that enable conversational in-car experiences, voice biometrics, navigation, and content access. Cerence serves global car manufacturers, tier-one suppliers, and mobility service providers, offering scalable platforms that integrate with vehicle infotainment systems and digital cockpits. Its portfolio spans speech recognition, edge AI, and cloud services designed to enhance driver safety and user experience across passenger vehicles, two-wheelers, and commercial fleets.
Subscribe to the weekly newsletter for similar remote roles and curated hiring updates.
Newsletter
Weekly remote jobs and featured talent.
No spam. Only curated remote roles and product updates. You can unsubscribe anytime.
Similar Opportunities

Trustly Group AB
2 months ago

Trustly Group AB
2 months ago

Ruby Labs Ltd.
2 months ago

Rerun Technologies Inc.
2 months ago