AI Research Scientist: Next-Gen Language Models & Agents

AMD; Santa Clara (CA); On-site; USD 180,000 - 260,000; Full time

Job summary

AMD is seeking an AI Research Scientist in Santa Clara to advance language models and intelligent agents for complex technical tasks. You will explore how models acquire domain knowledge, reason, use tools, and learn from real development environments across software and hardware contexts.

The role spans model training, agent design, and rigorous evaluation to deliver advances that improve software and hardware development quality and productivity. Strong Python and DL expertise required.

Qualifications

• Master’s or PhD in Computer Science, ML, Electrical Engineering or related field.
• Publication record in ML/AI venues is preferred.
• Experience with LLM pre-training, continued pre-training or fine-tuning and reinforcement learning.

Responsibilities

• Research methods for reasoning, planning, tool use, memory, and reliable execution of multistep tasks.
• Develop pre-training and post-training approaches to strengthen model knowledge and coding abilities.
• Build data pipelines using curated corpora, demonstrations, synthetic data, and agent trajectories.
• Apply reinforcement learning and adaptation methods to improve agent performance in executable environments.
• Investigate inference-time reasoning and search to improve solution quality and efficiency.
• Develop agents for software and hardware workflows including code generation and verification.
• Create benchmarks for correctness, generalization, and sustained performance across projects.
• Collaborate with software and hardware teams to translate research into capabilities.
• Publish research findings through technical reports and working demonstrations.

Skills

Python

Deep Learning

Language Models

Reinforcement Learning

LLM

Research

Distributed Training

Education

Master’s/PhD in CS/ML/EE

Tools

PyTorch

TensorFlow

GPU kernels

Job description

AMD is seeking an AI Research Scientist in Santa Clara to advance language models and intelligent agents for complex technical tasks. You will explore how models acquire domain knowledge, reason, use tools, and learn from real development environments across software and hardware contexts.

The role spans model training, agent design, and rigorous evaluation to deliver advances that improve software and hardware development quality and productivity. Strong Python and DL expertise required.