Autonomous AI Agent Engineer – LLM & RLHF Expert

Ethereum Technologies; Sunnyvale (CA); On-site; USD 77,000 - 110,000; Full time

Job summary

Ethereum Technologies seeks an Agentic AI Engineer in Sunnyvale, CA to design and implement autonomous AI agents and robust evaluation frameworks. You will analyze LLM performance, apply tuning techniques, and collaborate with researchers to push the boundaries of agentic AI.

The role emphasizes practical AI system design, scalable deployment, and staying current with advances in LLMs and reinforcement learning.

Qualifications

• Master's or Ph.D. in Computer Science, Machine Learning, AI, or related field.
• Strong foundation in AI/ML/deep learning principles.
• Hands-on experience building AI agents or robotics systems.

Responsibilities

• Design, develop, and implement behavioral logic and decision-making for autonomous AI agents.
• Develop evaluation metrics and testing frameworks to assess agent performance.
• Analyze LLM performance and identify improvements within agentic contexts.
• Apply LLM tuning techniques (fine-tuning, prompt engineering, RLHF).
• Collaborate with researchers and engineers to explore novel AI approaches.
• Contribute to AI platform and infrastructure to deploy scalable agents.
• Stay updated on AI/ML advancements in agentic systems and LLMs.
• Document design specs, implementation details, and experimental results.

Skills

Python

LLMs

Reinforcement learning

RLHF

AI agent systems

Debugging

Communication

Education

Master's or Ph.D. in CS/ML/AI

Tools

Hugging Face Transformers

PyTorch

TensorFlow

Vertex AI

Google Gemini

Job description

Ethereum Technologies seeks an Agentic AI Engineer in Sunnyvale, CA to design and implement autonomous AI agents and robust evaluation frameworks. You will analyze LLM performance, apply tuning techniques, and collaborate with researchers to push the boundaries of agentic AI.

The role emphasizes practical AI system design, scalable deployment, and staying current with advances in LLMs and reinforcement learning.