Job Description
About the Opportunity
Our client is a leading autonomous driving technology company with in-house silicon and massive-scale compute infrastructure. Their Agentic Lab is hiring a Research Scientist / Research Engineer to lead AutoResearch — using automated systems and AI agents to autonomously discover, optimize, and train next-generation LLM architectures, co-designed with their proprietary chip platform.
You will bridge hardware constraints (silicon architecture, memory bandwidth, latency, power) and model design, ensuring automatically discovered architectures match or exceed the pre-training efficiency and downstream performance of state-of-the-art base models.
What You'll Do
- Design, build, and scale automated search and optimization systems (NAS, evolutionary algorithms, LLM-driven research agents) to discover optimal LLM architectures
- Integrate hardware-level constraints (tensor parallel limits, memory bandwid...