Description
We are seeking a high-caliber Deep Learning Engineer to bridge the gap between cutting-edge multimodal architectures and real-time robotic execution for autonomous vehicles. In this role, you will design and implement SOTA algorithms to make LLM/VLM fast, lean, and reliable enough to power an end-to-end driving stack.
Your responsibilities will include:
Developing SOTA model optimization techniques, such as speculative decoding with block diffusion, KV cache streaming, and Prefill–Decode separation, etc. to boost E2E model performance for production deployments. Implementing advanced compression techniques including Quantization (FP4/FP8), pruning, and knowledge distillation to minimize model footprints without compromising safety-critical accuracy. Designing high-performance optimization strategies for inference, including automated model sharding (tensor/sequence parallelism) and the development of efficient attention kernels optimized for KV-caching. Conducting deep, layer-by-layer model profiling to identify compute and memory bottlenecks, driving targeted optimizations for real-time execution. Leveraging the PyTorch ecosystem to extract standardized model graph representations and automate deployment pipelines for TensorRT conversion. Scaling DL model performance across diverse NVIDIA edge architectures, maximizing the throughput of specialized accelerators on the road. Architecting the software interface to seamlessly integrate and interact with large-scale models within a high-performance C++ production environment. Partnering with research, TensorRT, and Cosmos teams to translate breakthrough innovations into shipping product solutions.
You will be working on a team dedicated to making self-driving vehicles a reality and believe this technology can save millions of lives. Join a team of innovative thinkers at one of the world's most respected technology companies.