Applied AI Researcher: Production-Ready LLMs and RL

Cisco Systems, Inc.; San Jose (CA); On-site; USD 216,000 - 281,000; Full time

Job summary

Cisco Systems, Inc. is seeking an Applied AI Researcher to join its AI Platform Team and translate state-of-the-art AI research into production systems that power Cisco's enterprise ecosystem.

You will work with AI researchers, data scientists, and engineers to deploy scalable, secure AI solutions across Cisco's ecosystem. You will advance LLMs and agent-based models, design evaluation methods, and drive production deployments across multiple domains.

Qualifications

• PhD with 5+ years in applied AI/ML research or engineering.
• Experience designing, training, and fine-tuning language models for production.
• Proficiency with Python and deep learning frameworks (PyTorch, TensorFlow, JAX, Transformers).
• Experience with post-training methods such as LoRA, QLoRA, PEFT, instruction tuning, or RL from human/AI feedback.
• Track record leading applied AI projects from research to deployment.

Responsibilities

• Collaborate with AI researchers to translate methods into production models on NeuralFabric.
• Design, train, and fine-tune LLMs for network, security, observability, and operations use cases.
• Advance RL and agent-simulation methods and create reusable training/evaluation recipes.
• Define evaluation methods and gates to ensure measurable, repeatable model quality.

Skills

Python

PyTorch

TensorFlow

JAX

Hugging Face Transformers

LoRA/PEFT

RLHF

Research leadership

Education

PhD in Computer Science / ML / AI

Tools

Auto Gym

Gym environments

Job description

Cisco Systems, Inc. is seeking an Applied AI Researcher to join its AI Platform Team and translate state-of-the-art AI research into production systems that power Cisco's enterprise ecosystem.

You will work with AI researchers, data scientists, and engineers to deploy scalable, secure AI solutions across Cisco's ecosystem. You will advance LLMs and agent-based models, design evaluation methods, and drive production deployments across multiple domains.